What Went Wrong:

Lessons from Manila, Buenos Aires, and Atlanta

By Kathleen Slattery
Director - Water, Sanitation, and Urban Services Practice
The Institute for Public-Private Partnerships (IP3), Inc.

About the Author...

Kathleen Slattery manages the Water, Sanitation, and Urban Services Practice of IP3. She teaches and consults on a wide range of PPP and regulatory topics, including institutional issues, procurement and contracts, and stakeholder consultation.

Introduction

The last year has been a difficult one for public-private partnerships (PPPs) in the water sector. While proponents of PPPs see them as valuable tools for improving the efficiency of water services and attracting capital for investment in system improvement and expansion, detractors cite private operators as profiteers that will let the quality of service deteriorate while increasing the price for water. The cancellation of several high profile PPP arrangements has added fuel to the fire of what is already widespread distrust in the concept of private involvement in the provision of water services.

PPP in the water sector is the subject of much debate at the Third World Water Forum in Kyoto, Japan. As water experts, governments, private operators, and civil society representatives from around the world gather to debate the merits of private involvement in the delivery of water services, this article examines major issues in three recent, high-profile water PPP arrangements, and highlights the lessons to be learned for Governments considering entering into PPPs.

Manila

In 1997, the Government of the Philippines entered into two PPP arrangements for the provision of water services in Manila. A 25-year concession agreement was signed between the Metropolitan Waterworks and Sewerage System (MWSS) and Maynilad Water Services, Inc. (MWSI) for service to approximately six million consumers in the western area of Manila¹. Maynilad is 59% owned by the Philippine-based Lopez family's Benpres Holdings (which also owns Manila Electric Co.) and 40% owned by the French water giant Ondeo.

The agreement was terminated on February 7, 2003 and responsibility for water services reverted to MWSS. Per the terms of the concession agreement, the termination procedure was initiated on December 9, 2002 when Maynilad issued a "Notice of Early Termination" to MWSS, citing MWSS' serious breach of its obligations under the contract and giving 60 days' notice.

Although Maynilad cited numerous factors behind its decisions to terminate the contract, including severe droughts caused by the El Niño weather phenomenon, the effects of the 1997 Asian financial crisis, and delays in the completion of a river basin project, the primary reason for termination was MWSS' refusal to implement a rate adjustment that would have enabled Maynilad to recover large foreign exchange losses. Maynilad's bid for the concession, which was mounted just prior to the 1997 Asian financial crisis, was based on the assumption that the peso would depreciate no more than 2 to 3%. To compensate for any significant change in exchange rates, the bid included a provision for the recovery of foreign exchange losses throughout the term of the concession via an extraordinary price adjustment mechanism.

Prior to termination, Maynilad's investors had contributed US $143 million in equity and guaranteed a US $165 million bridge loan, in addition to assuming MWSS foreign and domestic debts amounting to US $800 million. During the first five years of operations, Maynilad realized losses of 4 to 5 billion pesos (approximately US $75 to 95 million), with actual annual revenues just enough to service the inherited MWSS debt. When the Asian financial crisis resulted in depreciation of the peso from P26: $1 to P53: $1² , Maynilad's debts effectively doubled and the company was no longer able to sustain its operating losses.

The privatization of Manila's water system in 1997 was considered the largest PPP arrangement of its kind to-date. The termination is the culmination of 6 tumultuous years in which Maynilad was faced with protests demanding water connections, opposition to rate increases and rate re-basing, and claims of poor customer service. Despite these difficulties, Maynilad achieved some significant improvements in service, including increasing coverage from 58% to 84%. However, when compared to the positive experience of Manila Water, the concessionaire serving the eastern zone of Manila, Maynilad's poor performance raises a number of questions about what went wrong. Was MWSS' refusal to increase rates to blame? Was it poor management of the business by Maynilad? Or did extraneous factors such as El Niño and the Asian financial crisis create an environment in which the concession was doomed to failure?

The answer is likely a combination of factors. At the time the concession was awarded, no one - including the Philippine Government, its advisors at the International Finance Corporation (IFC), and bidders such as Maynilad - could foresee the effects of the impending Asian financial crisis. As a result, the terms of the concession agreement were drafted such that the impact of ordinary currency fluctuations - on the order of the 2 to 3% predicted by Maynilad - would be treated as "pass throughs" in the tariff, with tariff increases phased in over the life of the concession. This provision significantly reduced the risk to the concessionaire and made the transaction more attractive given the large amount of foreign debt to be assumed from MWSS. While the risk to the concessionaire was reduced, the risk to consumers increased, as they would be expected to bear the burden of higher tariffs. If currency fluctuations had been of the magnitude predicted by Maynilad, the corresponding tariff increases would have been gradual and reasonable. Instead, when dramatic devaluations resulted in the need for larger tariff increased, the regulator was forced to balance Maynilad's financial requirements with consumers' affordability.

This dilemma highlights the difficulty of developing a long term contractual mechanism such as a concession that is rigid enough to reduce each party's exposure to risk, while remaining flexible enough to adjust to changing circumstances over what are often 20 to 30 year contract periods. The irony in this case is that the changing circumstances occurred just after contract award and not 20 to 25 years in to the arrangement.

In addition its financial difficulties, Maynilad was unable to achieve operating efficiencies equivalent to those of its rival Manila Water in the eastern zone. By the fourth year of operations, Maynilad's operating expenses were a third higher than the company had projected in its bid, whereas Manila Water's were a quarter lower than it had projected. At the same time, Maynilad's cost per cubic meter to produce and sell water (calculated by dividing revenues by billed daily volume multiplied by 365 days) was P9.06 per cubic meter, as opposed to Manila Water's P4.71 cost per cubic meter. Arguably, the operating environments and therefore the investment requirements in the two concession areas were markedly different, necessitating differing cost structures. However, as one Philippine journalist put it, "It takes Maynilad almost twice as much as the competition to produce and sell a cubic meter of water."³

Another factor cited in Maynilad's failure is the Philippine Government's constitutional limit on majority foreign ownership of public utilities. Because majority ownership must rest in Philippine hands, PPP arrangements must be awarded to consortia that are majority locally owned. This presumes that there are local firms capable of taking on the financial burden (including operating expenses, debt service, and funds for new investments) associated with a water utility the size of Maynilad, and that such firms will remain viable over the life of the arrangement. In the case of Maynilad, when Benpres Holdings was unable to sustain operating losses and service its foreign debt, its partner Ondeo was limited in its ability to step in by its 40% ownership stake (the maximum amount of foreign ownership allowed by law). To address just this issue, the Philippine Chamber of Commerce and Industry (PCCI) has specifically requested that the Government consider a constitutional amendment to remove the limit on foreign ownership of utilities.

The experience with PPP in Manila's west zone has still not ended. Maynilad and MWSS are in arbitration proceedings to determine exactly how and under what terms the contract will terminate. Meanwhile, Manila Water continues to provide service in the east zone of Manila under its long-term concession.

Buenos Aires

In May 1993, the Government of Argentina awarded a 30-year concession for the provision of water and sanitation services in the City of Buenos Aires to the Aguas Argentinas consortium. Aguas Argentinas is 46% owned by the French firm Ondeo (with the Spanish firm Aguas de Barcelona and the UK's Anglian water holding minority shares), 37% Argentine-owned, and 10% employee-owned. Under the arrangement, Aguas Argentinas provides drinking water to 7.6 million residents and sewerage services to 5.7 million residents in the Greater Buenos Aires area.

Since the concession was awarded, the experience has been mixed. Tariffs were increased continuously, there were frequent contract renegotiations, several projected investments were cancelled, and the regulator's capacity to assume responsibility was questioned. However, in the ten years since the contract was awarded, Aguas Argentinas has increased water coverage from 70 to 83%, collection efficiency was increased from 90% to 95%, and operating costs as a percent of revenues dropped from 99% to 61%.

Prior to privatization, water prices were increased by 25% in February 1991, 29% in April 1991, and 8% in May 1993. The Aguas Argentinas consortium offered the lowest bid and was awarded the concession based on its promise to reduce tariffs by 26.9%. As was the case in Manila, the concession included a provision allowing for extraordinary price adjustments, but only in the first ten years of the concession and only in the case of increases of more than 7% in the price index contained in the contract or where unforeseen events that are beyond the concessionaire's control have occurred.

The extraordinary adjustment mechanism was first applied eight months into the concession, when Aguas Argentinas requested a tariff increase due to unexpected operational losses due to inadequate consumer records and the poor condition of the pipe network. As a result, tariffs were increased by 13%. Since that time, there have been further needs for increases and each time the Government and the private operator, with the intervention of the regulator, have been able to reach a workable solution.

In late 2001 and early 2002, the Argentine economy came crashing down and a financial crisis ensued when the peso was de-pegged from the dollar. The combined effect of the pesofication of tariffs and the devaluation of the peso in 2002 was a drastic decrease in Aguas Argentinas' ability to service its mostly U.S. dollar denominated debt.

In response to the financial crisis, the Government of Argentina passed the Argetinean Public Urgency legislation, which included a provision that enabled Aguas Argentinas to suspend repayment of its debts with multilateral finance institutions in order to redirect those funds towards ongoing operations. In response, the Argentine water regulator ETOSS is considering granting Aguas Argentinas a 15% tariff increase.

The outcome of the financial crisis and its ultimate impact on the success of the PPP arrangement in Argentina is still unknown. Recently, Ondeo released a statement saying that it was invoking the terms of the French-Argentine bilateral agreements that required the GOA to compensate French firms for any investments they have made (in other words, Ondeo is asking for either a tariff increase or some other recourse to collect the funds to cover its investments-to-date plus a reasonable return).

Atlanta

In January 1999, the City of Atlanta awarded a much-heralded 20-year management contract to United Water. The form of contract employed in Atlanta - an "Operations and Maintenance Agreement" or management contract - is typically used where the primary goal of the contracting party is to improve the quality of management and realize efficiency improvements in the provision of service. Often, the main "efficiency improvement" is actually cost savings achieved through increased revenues and decreased operating expenses. The contract in Atlanta is unusual in that its duration is 20 years. The typical management contract lasts from three to eight years and is then re-bid in order to realize the savings that can be achieved through a competitive bidding process.

Prior to entering into this PPP arrangement, water services in Atlanta were in need of efficiency improvements on a number of fronts. The EPA was fining the City for violating clean water standards, and the water department was seriously overstaffed (by 50% according to many estimates). In order to address the problems, the water department was proposing to double the tariff.

In Atlanta, as in Manila and Buenos Aires, the experience with privatization has been mixed. The utility has been brought into compliance with the EPA and is no longer paying fines. Audits have shown that United Water was saving the City of Atlanta $10 million a year. However, this is approximately half of the savings that United Water predicted in its bid. And while United Water's performance in some cases was not in conformance with contractual requirements, it was still outperforming the previously publicly run operation. For example, while the contract called for 4,500 fire hydrants to be repaired annually, United repaired approximately 4,000 - which is actually 1,000 more than the City repaired in the year prior to contract award. All of this has occurred with only a 10% increase in tariffs (as opposed to the 50% proposed by the water department prior to PSP).

Just four years into the 20-year arrangement, the City of Atlanta and United Water agreed to terminate the contract. One of the reasons that United Water cites for its mixed performance is the lack or inaccuracy of system data available at the time of the bid. To measure the contractor's progress under a management contract, the contracting authority must have well-defined baseline data that can be used to judge incremental improvements in performance. In the case of Atlanta, such data was in many cases either not available or inaccurate. All of the firms that bid on the contract in Atlanta were aware, prior to submitting their bids, of the deficiencies in the data. Arguably, United Water should have factored this in to its bid.

Was the lack of data on the Atlanta water system to blame for the termination of this arrangement? Was it United Water's failure to perform in accordance with contractual requirements? Or were there other factors at play? Again, the real reasons the Atlanta contract was cancelled are probably a combination of all three factors.

There are three primary reasons why most management contracts fail. They are: 1) lack of baseline data on system performance results in the establishment of performance targets that are unrealistic or unachievable; 2) the management contractor is not given full legal authority to carry out its mandate; and 3) political support for the arrangement wanes or political interference prevents the contractor from meeting its targets. In Atlanta, all three of these factors led to the contract's termination.

Independent audits carried out as part of the termination proceedings have supported United Water's claim that much of the baseline data was inaccurate. As a result, many of the targets that United Water was expected to meet were set at unrealistic levels. When the company has failed to meet them, it is penalized under the contract and both public and political opinions of United Water's performance are tarnished. In most management contracts, when it is discovered that baseline data was inaccurate, targets are reset or renegotiated. In Atlanta, the City denied that there were any problems with the data and refused to consider resetting or renegotiating tariffs.

Another reason cited for United Water's poor performance was its lack of legal authority to fulfill the terms of the contract. For example, one of the City of Atlanta's primary complaints related to United Water's performance in meeting billings and collection targets. While the company was expected to achieve a collection rate of 95%, it was only collecting 90% of its bills. United Water has offered two explanations for this shortfall. The first is that it lacks the legal authority to shut off service to customers not living within city limits and therefore in the county's jurisdiction. The second is that under the previous administration of Mayor Campbell, the Mayor's office had prohibited United Water from cutting off service to certain customers.

While political interference was a problem, so was the change in political support for the contract arrangement. Current Mayor Shirley Franklin's administration is much less supportive of the deal than former mayor Campbell's administration. As a result, the City's willingness to make the relationship with United Water work soured, and communication and cooperation between the contracting parties have deteriorated.

In the end, the City of Atlanta and United Water agreed to "amicably" terminate the PPP arrangement, and the City has resumed control of water services and announced plans to raise tariffs and increase staffing in the water department.

Conclusion

Given the experiences in Manila, Buenos Aires, and Atlanta, what are the lessons to be learned and the conclusions to be drawn?

According to the World Bank, which recently undertook a study of the 2,500 financed infrastructure projects that it supported during the period 1990 to 2001, 3.5% (or 7 out of 202) of all cancelled contracts have been in the water sector. When measured in terms of investment value, the water sector represents 11.3% of cancellations (representing approximately 2001 US$ 4.5 billion). This implies that the water contracts that have failed were for the most part those with very large investment requirements - which generally mean concession arrangements (divestitures can easily be eliminated as they are almost unheard of in the water sector, and leases typically do not entail major investment on the part of the private operator).

This raises the question of whether governments are entering into concessions too hastily, particularly in countries where the poverty (and thus affordability) is an issue or where the macroeconomic environment has not been stable long enough to enable private operators to enter into 20- or 30-year arrangements with any degree of certainty about currency fluctuations. Throughout the 1990s, the concession model was seen as the answer to many countries' water problems. It turns the responsibility for service provision almost entirely over to the private sector (with only asset ownership remaining in public hands) and leverages large amounts of private finance for investments in increased coverage and system repair and rehabilitation. This makes concessions particularly appealing where coverage is low and pipe networks are in a serious state of disrepair. The problem is that these challenges are most prevalent in developing countries, where widespread poverty makes cost recovery a challenge.

According to the Bank, "the water sector contract cancellations were attributed to 'controversies over price increases', especially sparked by 'some consumers and politicians' and 'difficulties in collecting from consumers'." ⁴ Such problems are to be expected in the water sector, where most publicly-run water utilities, particularly in developing countries, set their tariffs artificially low and therefore are unable to collect enough to cover the cost of operations, maintenance, and depreciation. When PPP arrangements are introduced, the expectation is that the tariff will be set to ensure cost recovery, and as a result, the tariff often needs to be increased. This is particularly the case with concessions, which require the private operator to recoup its investments and a reasonable return through collection of the tariff.

In most cases where problems have arisen in PPP arrangements, contractual difficulties have been worked out through adjustment or renegotiation of key contract terms (in fact, more than 55% of water PPPs in Latin America were renegotiated during the 1990s). Such renegotiation requires good faith on the part of both the Government and the private operator. It also often requires Government to make decisions - such as agreeing to increase tariffs - that may be politically unpopular. Where a good working relationship and collaborative spirit are lacking, or where political motivations prevent Government from negotiating, contract termination is the result.

This highlights two key lessons that can be drawn from the experiences in Manila, Buenos Aires, and Atlanta:

Where affordability and/or macroeconomic stability are a concern, consider a step-wise approach to PPPs: After the initial rush during the 1990s to enter into concession arrangements, many donor agencies and Governments have begun to experiment with the concept of management contracting as a first step in the PPP continuum. Management contracts provide Governments o with an opportunity to bring private sector expertise to bear in improving utility management, improving the quality of system data, and achieving initial efficiency improvements. They can be valuable tools in preparing a water utility for a deeper level of PPP such as a lease or concession. Because they do not link the private operator's remuneration directly to the tariff, tariff increases can be phased in over time and cost recovery does not become a determining factor in the private operator's ability to fulfill its contractual obligations. Despite the benefits of management contracting, the approach does have significant drawbacks. In countries where water supply coverage is low and the existing network is in a state of serious disrepair, management contracting will not bring the capital required to make necessary investments in system expansion, repair, and rehabilitation. The key lesson for Governments is to fully explore all of the options for PPPs and to consider a phased approach if appropriate.

Establish a positive working relationship with the private operator and engage in regular communication: Often, PPP arrangements are seen as a panacea, and the expectations placed on the private operators are more than they can realistically achieve given the systems and institutions that they are inheriting. Governments should only enter into a PPP arrangement if they truly believe in the benefits to be gained. As a result, it is important that each party enter into the contract having already established a trusting relationship. This can be accomplished by ensuring during the contract negotiation stage that the contract itself embodies reasonable risk sharing arrangements, and a fair balance of penalties and rewards. If both parties remain committed to the concept of PPP and to preserving their positive relationship through effective communication, contract renegotiations or adjustments will be accomplished in a way that satisfies not only the Government and the private operator, but the consumer as well.

Despite the three high profile failures outlined in this article, most PPP arrangements - including those in the water sector - do not result in contract termination. The reasons for this are twofold - governments in many cases realize that the alternative to PPPs (publicly-provided water) will be less efficient and perhaps more costly, and both governments and private operators have a sufficient stake in the success of PPP arrangements that they will do everything in their power to make them work rather than walk away. In the end, however, the success of PPPs will be measured in terms of consumer satisfaction, with the best arrangements resulting in the delivery of safe water to more people at affordable rates.

¹ The western zone includes parts of Makati, Manila and Marikina, as well as Valenzuela, Kalookan, Navotas, Malabon, Las Pinas, Paranaque, Pasay and Muntinlupa and certain areas within Cavite Province.

²The exchange rate in effect in December, 2002, at the time the notice of early termination was filed.

³ Nery, John. "It Wasn't a Weak Peso that Sank Maynilad," Philippine Daily Inquirer (December 13, 2002), Section 1.

⁴Harris, C. ... [et al.] (2003). "Infrastructure Projects: A Review of Canceled Private Projects," Public Policy for the Private Sector, Vol. 252, (Washington, DC, USA), World Bank. 6 p.