Responsive regulation for water PPP: Balancing commitment and adaptability in the face of uncertainty

Abstract

Public–private partnerships (PPPs) can potentially overcome financial, quality, and efficiency issues common to public service provision in developing cities. The management of long-term infrastructure PPP requires complex administrative and regulatory inputs and sound institutional underpinnings to ensure partners’ commitment to the contract over long time frames. Additionally, PPP governance typically necessitates adaptation mechanisms in order to accommodate high degrees of uncertainty and deal with changing operational conditions and new information. This paper proposes that PPP performance – both relational and instrumental – is contingent on a delicate, dynamic balance of credible commitment and adaptive capacity. Comparative case analysis offers a dynamic picture of how rules, relationships, and the quality of regulation shape a PPP's capacity to adapt to external shocks and new information. The cases also demonstrate the tradeoffs of promoting fidelity to the partnership agreement and making needed adjustments to contract settings and further show that PPP resilience may be positive or negative. For example, high degrees of uncertainty in the operating environment can lock in the PPP decision and, thus, promote survival at the cost of instrumental performance.

Keywords:

• Water supply PPP
• Regulation
• Water supply regulation
• Responsive regulation

1 Introduction

Public–private partnership (PPP) in infrastructure provision is often employed to deal with financial, quality, and efficiency problems associated with government provision. The common proposition is that PPPs leverage strengths of the public (e.g., accountability, stability) and private sectors (e.g., innovation, efficiency, competition) and compensate for their respective weaknesses. But successful implementation requires complex administrative capacities and sound institutional underpinnings to ensure commitment over long time periods, during which initial conditions, internal dynamics and capacities, and external pressures change. As such, adaptability is also critical. The challenge is to design systems that promote responsiveness while simultaneously upholding commitment.

This paper posits that water PPP success is contingent on a delicate, dynamic balance of credible commitment and adaptive capacity secured via ‘responsive regulation’. While the former safeguards the contractual agreement, adaptability is need ed to accommodate the uncertainties and risks inherent to long-term infrastructure PPPs. Amongst others, water PPP risks are associated with factors that could affect supply of raw water (such as extreme weather events or environmental damage); costs of supply (e.g., economic crises or changes in input costs); or pricing of water, such as public opposition or political shifts that influence tariff policies. Other uncertainties include potential shifts in consumer demand or political events that interrupt operations or the flow of payments. The balance between commitment and adaptability in the face of these uncertainties depends on reliable information and fair processes to make needed adjustments to operating and investment plans – and even the agreement itself.

Comparative analysis of two Southeast Asian cities – Manila and Jakarta – offers insight into how rules and relationships can limit a water PPP's capacity to adapt in the face of uncertainty. The cases also demonstrate the tradeoffs of adaptability and fidelity to the partnership and show that PPP resilience can promote survival, but sometimes at the cost of performance. We begin with an overview of water PPP and its complexities, followed by a discussion of regulation as a key mechanism for balancing commitment and adaptability. Following case descriptions, we conclude with analysis and lessons for water PPP regulation.

The study is relevant to the regulation, contract, and institutional economic academic literatures, and also has important practical implications for water PPP, particularly in developing regions where uncertainties are often highest. The high costs associated with cancelations, non-adjustment of inefficient arrangements, and extended renegotiations drive demand for release valves to deal with systemic pressure. But these mechanisms must not be so flexible so as to incentivize rent seeking within the limits of the partnership agreement.

2 PPP in water supply

Traditionally, governments have been tasked with providing water services due to market failures, including monopoly and externalities, and to ensure access to a life-essential resource. Over the past twenty-five years, in response to perceived government failures including inefficiencies, rent seeking, and over-subsidization, market mechanisms have been employed along a spectrum of corporatization to full privatization. The World Bank Private Participation in Infrastructure (PPI) Database shows that, as of Q1 2016, 906 water and sanitation projects had reached financial closure, with total committed investments of US$79.6 billion. Of these, 63 projects were classified as either failed or distressed, representing 26% of total committed investments in the sector (PPI Database, 2016).

PPPs have been expected to introduce competition and knowledge to promote innovation and technical and efficiency gains (Bull & McNeill, 2007). These reforms were expected to “unleash the efficiency of the private sector” while simultaneously safeguarding public interest (Lobina, 2005). But performance outcomes have been mixed: some posit direct efficiency and performance gains (Estache, 2006; Galiani, Gonzalez Rozada, & Schargrodsky, 2009; Marin, 2009), while others cite examples of worsened performance (Prasad, 2006). The introduction of private capital is expected to relieve overburdened governments of financial shortfalls, though some caution that these arrangements straddle weaker governments with excess risk and long-term, inflexible financial obligations (Bloomfield, 2006; Flinders, 2005).

Performance outcomes and contract survival are dependent on both the partnership's fitness-for-purpose, as well as the ability for parties to successfully manage risks inherent to water PPP. These may include project-specific risks, such as construction, commercial, demand, and operating risk, or they may be external, including environmental, political, regulatory, currency, or force majeure risks.¹ A number of mechanisms may be employed to deal with these risks, usually according to the rule-of-thumb of allocating risks to the parties “best able to anticipate or respond to the risk factor” (Irwin, 2007).

Risk allocation is governed by specific service obligations; the structure of remuneration and pricing; and express contractual provisions that adjust the risk allocation implicit in the PPP's basic structure (Grimsey & Lewis, 2004). Specific tools include the contract itself, which codes each parties’ commitments; government guarantees to protect operators from demand risks or ensure a base level of returns; traditional insurance for extreme events; and structured mechanisms to allow adjustments to new information or external shifts, including scheduled reviews or a system to govern discretionary regulation. Adjustments are often made under the oversight of a regulatory body (or an agency mandated to oversee the agreement) tasked with overseeing a contract and/or imposing discretionary regulation based on professionalism, legal authority, and good information.

Even with these mechanisms for risk management and adjustment, the balance of multiple failures can tip back toward the market, leading to contract termination and reversion to government provision. A number of high profile cancelations in the 1990s and early 2000s cooled enthusiasm for water PPP, and recent debate has tended to portray it as a failed experiment (Araral, 2009; Bakker, 2010; Hall & Lobina, 2006; Hukka & Katko, 2003; Lobina, 2005). Indeed, advice regarding the suitability of PPPs to water services remains polemic. We have yet to definitively answer when water PPPs “work”. This paper argues that water PPPs can significantly contribute to water services, but only when the project architecture, adaptive capacity, and regulation allows for both systemic responsiveness and participant adherence to consensual goals in the face of uncertainty.

In linking adaptability and commitment to performance, it becomes important to specify what is meant by “success” or “failure”. In the case of PPPs, judgment may be made along instrumental-technical and relational dimensions. We use “instrumental performance” to describe how well a PPP attains the goals that motivated its implementation. Its assessment lies within the realm of traditional policy evaluation and typically involves measures such as service coverage expansion, reduction of non-revenue water (NRW), and cost reduction. “Institutional performance” instead refers to the degree to which parties cooperate, co-manage interests, remain committed to agreements, and sustain political support.

The clearest evidence of failure is contract termination, but this is only one manifestation. Indeed, a PPP may last despite poor technical performance if the contract is beneficial to powerful parties. In this case, ‘institutional failure’ is at hand, even if cancelation does not occur. Alternatively, a privately managed utility might attain performance goals and remain financially solvent but be terminated due to political opposition or infighting. These cases may be considered ‘institutional failures’. Projects that are generally successful or failed across both dimensions are systemic successes and failures.

A note on contract renegotiation is germane, as there has been much attention on the often tenuous renegotiation of PPPs (Hall, Lobina, & Corral, 2010; Ross & Yan, 2015), which may indicate failure, but may also be erroneously cast as such. While renegotiations can cause problems, they are also typical in long-term arrangements and allow for adjustments to external change or new information (Delmon, 2011). Indeed, the ability for parties to modify contract terms in changed circumstances can be important to systemic success over the long term.

The successful application of PPP to water is complicated due to inherent features that increase uncertainties. Reforms are complicated by the fact that water is “essential to life, local in supply, dull in innovation, and mysterious in information” (Shirley, 2007). Essentialness and symbolic weight mean that water is more politically contentious than other infrastructure services (Bull & McNeill, 2007). The local nature of water exacerbates this, since local politicians are responsive to intensely competing interests. These factors increase political risks of reform. Water is “dull”, in that supply is capital intensive with large fixed costs. This requires that prices be set so that small returns are accrued over long periods. As such, investments have lower and riskier returns over long timeframes, during which the policy environment may change. Finally, water is mysterious, in that subterranean networks are not easily visible, making assessment and valuation difficult (Shirley, 2007). Thus, the nature of water itself can magnify commitment and adjustment problems in PPP.

2.1 Responsive regulation: balancing credible commitment and adaptability

Regulation is a key mechanism applied to PPP arrangements to manage and allocate risks, hold partners accountable, and mediate adjustments to the partnership arrangement. It is important because the stability of a partnership and attainment of its aims depends on controlling opportunism and steering partners to uphold their ends of the bargain. Credible commitment reduces risks of future defection and, thus, ex ante and ex post transaction costs of partnership.

Shepsle proffers that commitments are credible in two ways: motivational and imperative. A commitment is motivationally credible if players want to honor the commitment, in which case the agreement is self-enforcing. A commitment is imperatively credible when the actor must honor the commitment because compliance is coerced (1991). Because participants juggle interests that shift over time, change in relative intensity, and interact with outcomes, imperative credibility is critical to ensure adherence to the agreement.

Regulation of water services typically includes oversight of pricing, the level and schedule of investments, quality standards and service obligations, and the rate of earned returns (Stern & Holder, 1999). The two archetypes are contract and discretionary regulation. The first sets specific terms of service and remuneration over a finite period, planning for a range of contingencies and relying on legal instruments. Contracts offer specific commitments and efficiency-inducing forces via competition for the market, but they are inevitably incomplete, and enforcement is dependent on the integrity of the judicial system. Discretionary regulation involves a government unit tasked with establishing tariffs and service standards open-endedly. The agency enjoys wide discretion under a framework of guiding principles, allowing increased flexibility but also increased risk of regulatory capture or mistakes due to information asymmetries (Gómez-Ibáñez, 2003). In developing regions, it is common to see hybrid models based on contract regulation but with regular adjustment periods where discretion is used (Wu & Leong, 2013).

Regulatory effectiveness is contingent on reduced information asymmetries; incentives motivating compliance; and supportive enforcement provisions (Shirley & Ménard, 2002). But monitoring, enforcing, and adjusting are generally complex technical activities that require technical capacity, regulatory independence, and administrative competence (Levy & Spiller, 1994). Clarity of roles and objectives, transparency, and predictability (Stern & Holder, 1999) are also critical underpinnings.

Central to good regulation are rules that protect against arbitrariness and opportunism and structure predictable processes, but less attention is given to the importance of responsiveness. This is important, because in many places where PPPs are employed, political and economic conditions are fluid, and information is limited, amplifying operational uncertainties and demanding adjustability. Furthermore, case studies of high profile cancelations and project distress in South America following the Argentine currency crisis and in Southeast Asia following the Asian Financial Crisis, as well as large-N quantitative research (House, 2013), shows that long-term water contracts are instrumentally and institutionally vulnerable to economic shocks. While extreme political resistance can also be the undoing of a PPP (e.g., Cochabamba), political volatility is not as significant to survival overall (House, 2013). One plausible explanation is that political volatility can prevent cancelation when government capacity to negotiate termination and manage re-municipalization is undermined due to political turmoil. This suggests that PPPs can remain institutionally steady in the face of political volatility but says nothing about effects on instrumental performance. In other words, a PPP that lasts may nonetheless not be adjusted sufficiently to be successful from a policy perspective.

A major challenge is that information required to forecast demand and costs is often unreliable (or unavailable). As such, the assumptions upon which contracts are built and tariffs determined are often uncertain. Many contract problems are rooted in overambitious and unworkable commitments that cause subsequent financial problems (Van den Berg, 2000). Other problems stem from ambiguously defined performance measurement terms and operating rules.

Thus, there is a need to balance accountability with flexibility to external change, internal feedback, and new information. In this vein, Selznick proposes the notion of responsive regulation, wherein the challenge is “to maintain institutional integrity while taking into account new problems, new forces in the environment, new demands, and expectations” (1994). Responsiveness is also a key to water PPP regulation, since participants have markedly different interests influenced by external commitments that shift over time.

Delivering responsiveness whilst preserving commitment is not easy, however. Agreements may be difficult to fairly adjust due to unspecific decision rules or weak appeals processes. Low technical capacity and institutional weaknesses exacerbate this in many developing regions, where many regulators are inexperienced and vulnerable to political interference (Tecco, 2008). Kettl points out that “government's relationships with the private sector are not self-administering: they require rather aggressive management by a strong, competent government” (1993). Lobina also argues that interactions between the government and PPP operators “are strongly affected by the unequal distribution of resources and skills between the parties” (2005). This can lead to governments being outdone with respect to negotiating contract terms.

Other factors are cultural and relational. In studying Latin American water concessions, Post shows that local operators are more successful due to greater flexibility, resilience, and knowledge of local conditions. Furthermore, they are more apt to engage in “relational contracts” that allow direct negotiation in the face of unforeseen circumstances (Post, 2009). House's study also shows that foreign participation increases the likelihood of cancelation (2013). Jacobsen and Choi identify other relational success factors, including joint commitment, open communication and trust, and willingness to compromise (2008). These all suggest that trust – more easily formed amongst participants with common cultures and mental models – is key to motivational credibility and is an important foundation for sustaining PPP through contentious processes of high uncertainty.

3 Water PPP in Southeast Asia: case study of Manila and Jakarta

Comparison of the water PPP experiences in two major Southeast Asian cities, Jakarta and Manila, allow us to examine the relational, contractual, regulatory, and information conditions that simultaneously foster adaptability and commitment. These cases have been selected for three reasons. First, they are suitable for most-similar comparative research due to similarities between the local operating conditions and the form of the PPP arrangements, which have nonetheless yielded different outcomes. Despite a contract termination in Manila, the city has emerged as an example of a productive PPP scheme. Jakarta, on the other hand, despite continuance of its concessions, is widely considered an instrumental failure. Second, both cases cover cities with hybrid regulatory approaches that employ both contract and the ongoing oversight of a regulatory body. This is important because (a) we can observe adjustments to contractual agreements that are made through a public forum (the regulatory body), and (b) hybrid forms are observable in other developing region contexts, improving the external validity of findings. Third, both cases have experience multiple external shocks (economic crisis, political shifts, etc.) over time, allowing for multiple observations of adjustment. Whilst that there are undoubtedly numerous paths to failure or success, there are general lessons to be drawn from comparing regulatory systems, relational conditions, and each city's ability to deal with uncertainties over time.

3.1 Manila

Manila privatized water services in 1998, via two concessions that represented the world's largest water PPP, with projected investments of US$7 billion and a service area supplying 11 million residents (Dumol, 2000). The PPPs were established to overhaul a poorly performing system managed by the Metropolitan Waterworks and Sewerage System (MWSS). By the mid-90s, the agency owned a deteriorated and leaking network, operated at huge financial losses, and depended heavily on subsidization (Wu, Batac, & Malaluan, 2011). MWSS served only two thirds of Manila residents with intermittent water, and less than 10% with sewerage.

In response to a projected water emergency, the 1995Water Crisis Act opened a one-year window to privatize. While rushed, the Ramos government facilitated a planning process that yielded specific and extensive contract terms and jump-started efficiency and cost recovery measures (Wu et al., 2011). To allow benchmarking and competition, the city was split into East and West service zones. Bid terms charged concessionaires with operational responsibility for 25 years and capital investments to expand services to full coverage and reach targets for service and NRW reduction. Operators would retain revenues, assume MWSS debt, and pay yearly concession fees to fund MWSS. The bid rules required that project companies comprise a local sponsor with majority control and a multinational with expertise (Wu et al., 2011).

In 1997, winners were selected based on lowest tariffs, which were to remain in place until the first rate rebasing in 2003. The East zone was awarded to Manila Water, a consortium of the Ayala Corporation, United Utilities, and Bechtel, who submitted a surprisingly low tariff of nearly a quarter of the pre-concession rate (Wu & Leong, 2013).² Philippines’ Benpres Holdings and French Suez Lyonnaise des Eaux was awarded the West under their joint venture, Maynilad. Both concessionaires’ low bid rates assumed significant and rapid efficiency gains.

Tariffs were to be adjusted regularly by automatic indexation for consumer price index shifts; annually for unexpected impacts via an Extraordinary Price Adjustment mechanism³; and during five-year rate rebasings. Rebasing would adjust tariffs to account for planned expenditures and reasonable returns, not to exceed an appropriate discount rate. A Regulatory Office (RO) was established within MWSS to oversee the contracts and facilitate rate adjustments. To speed PPP adoption and overcome legislative obstacles, the government established a “semi-autonomous” RO via contract, rather than an independent agency (Dumol, 2000),⁴ and an Appeals Panel with binding power to manage disputes.

The RO would oversee contract clauses on rates and monitor service commitments. In the beginning, their actions were largely limited to monitoring and routine tariff adjustments (Lazaro, 2002). The RO was afforded limited formal training and few resources to learn technical regulation (Interview, MWSS-RO, 2013).⁵ As such, while MWSS and the operators were contractually bound to adhere to decisions of the RO and Appeals Panel, the RO's degree of coercive power was uncertain.

Early 1998 brought the first regulatory challenge – and the first opportunity to establish public legitimacy. While low bid rates had made PPP politically saleable, operators soon discovered that they were overly optimistic. Drought limited the raw water supply, and the Asian Financial Crisis led to currency devaluation that nearly doubled debt burdens and grossly increased costs. In March, both concessionaires applied for an extraordinary price adjustment (EPA). Despite limited experience, the RO saw the event as an opportunity to set the tone for regulation.

The debate that followed centered on contract ambiguity over the appropriate discount rate (ADR) used as the upper limit on returns, which had not been specified by contract. The RO denied the Manila Water's request based on an 18% ADR and front-loaded loss-recovery, granting a much lower increase based on the 5.2% implied in the original bid. Manila Water appealed, and the Appeals Panel ruled in the operators’ favor.⁶ Nevertheless, the episode confirmed RO commitment to consumer protection and legitimized the body to operators and the public.

The validation of a strong RO would not serve Maynilad well, however, as it struggled to deal with shortfalls and strategic missteps.⁷ After the crisis, their debt virtually doubled, bringing debt service above revenues (Kumar & Brock, 2012). The company had also pursued an expensive expansion strategy that failed to attain efficiencies. By 2000, operating expenses were a third higher than bid projections, and their production cost was nearly twice Manila Water's (Finger, Allouche, & Luis-Manso, 2007). Parent company Benpres was also in financial duress, crippling Maynilad's borrowing capability. In October 2000, Maynilad requested an accelerated EPA to recover foreign exchange losses, but the RO granted only a modest increase. In March 2001, Maynilad declared force majeure and halted concession fee payments.

Under these disputatious circumstances, the RO initiated the first scheduled tariff rebasing in 2002. Aware of its own capacity limitations, the RO engaged an expert committee from the University of Philippines to evaluate rate inputs and expansion plans, determine the ADR, and recommend tariffs. The team developed a framework based on the ‘efficiency and prudence’ of expenditures and fairness of the ADR. The contracts lacked detail on these issues, but the team developed working tests and clarified ambiguities associated with service obligations, measurement, and information requirements (Interviews, MWSS-RO, 2013).

The consulting team recommended reductions in planned expenditures for both operators and “disallowances” of PHP8.8 billion for Maynilad and PHP643 million for Manila Water against their opening cash positions.⁸ The striking difference indicated a judgment of Maynilad's relative inefficiency. The team suggested a rebased rate of PHP24.00 for Maynilad against their requested PHP34.72, and PHP17.00 for Manila Water against a requested PHP19.54 (Wu et al., 2011). Manila Water accepted the outcomes, but Maynilad again disputed. ⁹

In response to noted challenges, including missing target measurement definitions and insufficient information, the RO and consultants developed a new monitoring and oversight program that required regular reporting of key performance indicators (KPI) and business efficiency measures (BEM) (Interviews, MWSS-RO and Millan, 2013).¹⁰ Rebasing also exposed strengths of the system, particularly the benefits of benchmarking and the RO's ability to keep concessionaires on task. This marked a shifting role for the RO from contract oversight to more discretionary action, both during rebasing and during the interim periods via the reporting system. Concessionaires also supported interim reporting, since it allowed incremental adjustments and minimized the likelihood of surprises during rebasing (Interviews, MWSS-RO; Millan, 2013).

Both operators made progress in service expansion and NRW reduction, but whereas Manila Water turned profits starting in 1999, Maynilad continued to suffer huge losses (Finger et al., 2007). In December 2002, Maynilad filed notice of contract termination followed by bankruptcy. MWSS began to orchestrate re-municipalization, initiating contract termination in 2005. MWSS assumed ownership with a debt-for-equity swap. In late 2006, after two years of municipal ownership, Maynilad was re-awarded to a joint venture of DMCI Holdings and Metro Pacific Investments Corporation. The operator became immediately profitable and made quick improvements to NRW, service, and coverage. By the mid 2000s, both operators were widely cited as water supply PPP successes.

Through crisis recovery, contract amendments, rebasing, and the Maynilad transition, the RO's role shifted toward discretionary regulation as the organization became regularly involved in assessing operations and investments and mediating contract adjustments. The introduction of a continuous monitoring system and clarifications to contract ambiguities smoothed the way for a relatively easy second rebasing in 2008 (Interview, Alikpala, 2013).

Maynilad expressed intentions to dispute RO tariff recommendations during the second rebasing, but the company was dissuaded when Manila Water orchestrated a valuable contract extension for both operators following a politically driven tariff freeze. Firms argued that it would be impossible to meet targets without extended terms. Against RO wishes, fifteen-year extensions were granted. The experience betrayed the RO's limited independence, but performance steadily improved, marked by significant reductions to NRW.¹¹

Not until a contentious tariff ruling in the third rebasing of 2013, however, would customers potentially recoup efficiency gains. Rising tariffs had invoked public opposition.¹² While increases were due in part to poor forecasting and over-confidence during bidding, they were also attributable to effective capital expenditures that brought about significant improvements. The third rebasing would nevertheless test the evolved RO.

Operators requested rate increases, but following an extensive audit, the RO recommended reductions to capture efficiency gains and bring rates under the threshold of willingness to pay (Interview, MWSS-RO, 2013). Lengthy board resolutions were published to explain decisions, which centered largely on technical calculations but also touched on interpretations of RO roles and responsibilities. ¹³ Resolutions stated, “the Regulatory Office has been mindful of its mandate to ensure adherence of the parties to the CA. The Regulatory Office is also keenly aware, however, that the CA does not exist in isolation – it must be viewed and interpreted alongside applicable provisions of law and in a manner that consistently underscores the Concessionaires’ duty to ‘comply with all the laws, statues, rules, regulations, orders and directives of any governmental authority that may affect the Concession from time to time… It bears stressing that the CA is no ordinary contract, but one imbued heavily with public interest.”

3.2 Jakarta

Indonesia's capital city boasts 10 million residents in its densely packed city proper and thrice that in the greater metropolitan area. As Indonesia's most important economic and political center, the city holds status as a special administrative district (DKI Jakarta) led by the Governor. Jakarta's infrastructure is critical to the country's industrial and economic growth, and water politics are intense and played out at multiple governmental levels.

Until PPP, a government-owned company, PAM Jaya, was responsible for supply.¹⁴ The agency had been unable to expand the network to serve the growing population and could not generate sufficient revenues for maintenance. By 1997, PAM Jaya was intermittently servicing 43% of the city with non-potable water. Production costs and tariffs were high compared to other cities in the region, and NRW had climbed to around 58% (Iwanami & Nickson, 2008). Poor services were attributed to low capacity, corruption, and mismanagement.

In the early 1990s, during the last years of Suharto's corrupt authoritarian regime, the government introduced an ambitious development program that included water privatization. The government quickly legalized private and foreign ownership of water supply and issued letters of invitation to two conglomerates for contracts to be directly awarded. The project companies were comprised of a multinational and a local company linked to the Suharto regime. UK's Thames Water joined Suharto's son to form Thames Pam Jaya (TPJ), and French multinational Suez linked up with a Suharto ally, Anthony Salim, in a joint venture called PT PAM Lyonnaise Jaya (Palyja).

PAM Jaya officials were blindsided – the agency was excluded from discussions as the companies negotiated directly with national government (Interview, Lanti, 2013). In mid-1997, two 25-year Cooperation Agreements (CAs) were signed. TPJ would cover the East and Palyja the West. The CAs stipulated goals of universal coverage, full cost recovery, and reduction of NRW to 20% by 2023. The companies were guaranteed a 22% internal rate of return over the contract period (Nugroho, 2011). PAM Jaya assets would be turned over to the concessionaires, and the companies took on seconded employees and assumed $231 million in foreign debt (Harsono, 2003).

The agreements stipulated a “dual tariff” instrument: the water charge owed by government to operators was separated from the tariff paid by customers. Concessionaires would collect payments and deposit them into escrow. Operators, in turn, would receive a flat rate per cubic meter of water. This was intended to balance revenues, given the socioeconomic differences between zones, and allow greater flexibility in levying tariff increases (Jensen, 2005). The Governor's office would set tariffs, but charges were based on a contracted formula.

Five-year water charge rebasing reviews would accommodate changed input costs, capital expenditures, and adjustments required to make guaranteed returns. When the contract was signed, there was an assumption that revenues would consistently exceed the water charge. This was not the case, however: the charge rose above the average tariff in late 1998, and remained so until 2003, creating serious financial problems for PAM Jaya (Interview, JWSRB 2012).

The CAs also exhibited ambiguities and design weaknesses that caused major information and authority problems. First, the contracts were limitedly performance-based, stipulating targets only for coverage and volume. Moreover, PAM Jaya was not granted access to financial or operational data, and there was no monitoring and oversight system in place (Lanti, Ali, Kretarto, Nugroho, & Zulfikar, 2009). With profits delinked from revenues and virtually no effective oversight, companies were not incentivized to improve efficiency or cost recovery. Moreover, while signatory parties would jointly decide investments, only operators could decide how to allocate them, giving wide latitude to change plans (Jensen, 2005).

The 1997 Asian Financial Crisis also rocked the Indonesian economy. Amidst upheaval in early 1998, the concessionaires took over. In the interim, the value of rupiah commitments had fallen from US$610 to $160 million. An angry public rebelled, bringing Suharto's resignation and the collapse of his longstanding regime in May. PAM Jaya attempted to terminate the concessions, but with help from their national governments, Suez and Thames convinced the new administration to uphold the contracts as long as Suharto-linked partners were dropped from the partnerships (Harsono, 2003). Ownership shifted to 95% stakes held by parent companies and 5% to other local contractors. The flashpoint proved surmountable, but the deals would limp along, damaged relationally, politically, and financially.

Participants entered a drawn-out, three-year contract renegotiation, intended to redress initial contract weaknesses. The companies were unhurried, however, as new contracts would likely reduce their benefits. Political resistance rose, and the Governor announced a tariff freeze until 2000 (Jensen, 2005). Tensions increased as PAM Jaya grappled with growing deficits and auditors attributed high costs to expatriate salaries and office rentals (Lanti et al., 2009) – a claim used to justify continued tariff freezes and intensified distrust of the foreign companies.

In October 2001, operators signed Restated Cooperative Agreements (RCAs) that improved some initial contract weaknesses. The most important change was establishment of a Jakarta Water Supply Regulatory Body (JWSRB) to mediate disputes and recommend tariffs to the Governor (water charges would remain formulaic). Also, whereas the original CAs had only stipulated targets for volume and water quality, added metrics included number of connections, production capacity, and NRW (Nugroho, 2011). PAM Jaya could levy penalties for missed targets. These changes did not, however, make a lasting difference: the RB was soon relegated to the background by stronger political actors, and the RCAs insufficiently specified measurement terms or details for monitoring and oversight, opening space for further dispute.

In this environment, the operators and PAM Jaya began rebasing for the second rate period set to commence in January 2003. Negotiations stretched to the end of 2004, and ultimately stalled in intense disputes over measurement issues, service and expansion priorities, and projected costs. The proliferation of clashes over nearly every input prevented progress, and the new RB struggled to find its place. Even government parties conflicted over their role: according to a former JWSRB member, DKI initially told the RB to “mind their own business”, and the Body did not know how to manage other government officials (Interview, Ali, 2013). Furthermore, the RB's influence was merely recommendatory, whereas the most important issues related to charges and performance remained the subject of negotiations between agency and operator.

During the protracted rebasing, the Governor introduced an independent consultant team to move rebasing along, which raised hopes for positive revisions. Unfortunately, both PAM Jaya and operators refused to cooperate, demonstrating a preference for internal dispute resolution (Jensen, 2005). In another round of RB-facilitated negotiations in 2004, operators and PAM Jaya agreed to an Automatic Tariff Adjustment (ATA) mechanism to semiannually adjust tariffs to cover water charges and PAM Jaya costs. It was also intended to cover a “To Be Determined” reserve component (Lanti et al., 2009) – yet another contract ambiguity and failure to establish precise, explicit, and enforceable terms. Further, while the ATA was by definition automatic, JWSRB asserted an obligation to ensure “that tariff adjustment awarded is not always as high as requested by the concessionaire” (Lanti et al., 2009), belying a non-technical approach to regulation.

Palyja finally signed a 2004 contract containing a revised base water charge, but TPJ failed to come to a new agreement. In the face of low performance, tariff freezes, and frustrated by the constant conflict, parent companies began to retreat. In 2006, Suez reduced its share of the west concession to 51%, and Thames exited completely, selling its 95% share to Indonesian Acuatico (Interview, Ali, 2013). The localized company changed its name from TPJ to Aetra, and the shift marked a new period for Jakarta water. Aetra began negotiations for a revised contract as the relationship between PAM Jaya and Palyja fell into a steady state of distrust, avoidance, and tit-for-tat legal action. As the parties moved toward the second rebasing, longstanding conflict had become the norm.

The second rebasing involved only PAM Jaya and the operators; JWSRB was altogether excluded. But once rates were agreed and presented to DKI, JWSRB blocked approval with an alternative recommendation. None of the parties reached a resolution, so the water charges set in 2003 were utilized by default, and performance targets were reduced. The only meaningful outcome was the imposition of a penalties ruling in 2008, which aimed to hold operators to task for missed 2004–2007 targets. Operators agreed to pay, but subsequent efforts to secure stronger oversight were hampered again by contract ambiguities, this time around the frequency of monitoring and penalty imposition.¹⁵ Persisting contract incompleteness left too much latitude (and too little guidance) on RB rulings, as well as too much space for operators with superior legal resources to argue around impositions.

In 2011, laws changed the role of the JWSRB, further subsuming the regulatory body under PAM Jaya.¹⁶ The weak regulator was virtually eliminated from management of the city's water services, leaving a vacuum in the mediation of disputes. A June 2011, letter from the Suez Chairman to the Indonesian Minister of Economic Affairs included a request for help with “serious difficulties” with PAM Jaya, which led to the untimely removal of the PAM Jaya director.¹⁷ Critics suggested that the threat of an increasingly powerful local water agency motivated companies to solicit high-level diplomatic support to remove him (Megarani, 2012).

In May 2012, Palyja presented its rebasing proposal for the third period. The following month, PAM Jaya signed a new Master Agreement with Aetra stipulating that PAM Jaya's shortfall would zero out at the close of the contract, tariffs would not increase, and the current water charge would only be increased at a fixed low annual rate for the remaining ten years (Interview, Kaderi and Limbong, 2013). Surprisingly, Palyja expressed little concern. As one manager commented, “It's all just a game – they will renegotiate in a year. The tariff has been frozen for the past seven years and is intended to stay for ten more. How can the system sustain seventeen years of frozen tariffs?” (Interview, 2013).

In step with a worldwide retreat from developing markets, Suez poised to exit Jakarta by mid-2012, announcing its intention to sell its stake to Manila Water. The sale fell through, and the government proposed a buyout of Suez’ stake. Whereas the Rp 330 billion shortfall with Aetra was cautiously expected to shrink to zero by 2016 (Interview, PAM Jaya, 2013), there was no plan to work down PAM Jaya's shortfall with Palyja of Rp 380 billion (Interview, PAM Jaya, 2013). The deals have left the local government with staggering debt, targets adjusted downwards to meet poor performance, and higher tariffs than most major Southeast Asian cities.

4 Uncertainty and ambiguity in water PPP: lessons from Jakarta and Manila

Several instances demonstrate the significance of uncertainty and ambiguity in Manila. Uncertainties surrounding operators’ ability to make rapid efficiency gains were negatively realized, initial bids were found to be unsustainably low, and the shock of the Asian Financial Crisis left companies saddled with excessive debts. Contract ambiguities and incompleteness were significant early challenges. The Jakarta concessions also suffered from incomplete contracts and the fallout of the crisis. But where Manila was able to adjust with the mediation of an increasingly adept RO committed to technical regulation and protecting public interest, Jakarta was left hobbled and unable to make needed adjustments following the external shocks and the discovery of inherent contractual problems.

The Manila PPPs experienced several significant adaptations facilitated by an increasingly discretionary RO. Operators secured tariff adjustments, revised targets for service expansion and investment, and secured contract extensions. While counter to program theory, they nevertheless allowed the concessions to recover from crises and government-side risks. They were legitimized by an RO noted for early consumer-oriented wins, transparency, and public consultation. Certainly, digressions from initial plans arose due to external shocks and commitment weaknesses on both sides. Whilst detrimental in some cases, they have not been ruinous as in Jakarta. This is largely because the participants have learned how to balance multiple interests and relieve pressures that may have otherwise led to termination or arrested later performance and efficiency gains.

The experience with Maynilad also shows that there is a limit to productive adaptiveness and a time for firm adherence to contract. Changes that are brought about by mutual agreement and publicly legitimized can accommodate learning and external shock, but those related to independent maneuvering to avoid commitments must be restricted. In Manila, the RO managed to hold concessionaires increasingly accountable to both the letter and spirit of the contracts while maintaining a reasonable cushion for adjustment. The RO demonstrated technical capacity improvements, continued efforts to redress contract ambiguities, and high levels of consultation, allowing for more palatable and effective adjustment to unexpected events. Growing technical capacity garnered the respect of key stakeholders, and published resolutions and active outreach clarified the RO's positions and guiding principles in discretionary rulings.

Jakarta, on the other hand, demonstrates that low adjustability can lead to institutional stasis alongside instrumental failure. Moreover, ambiguity can be an important tool to skirt adjustments that might otherwise upset a financially lucrative but societally unfavorable status quo. The rapid deterioration of commitments was motivational and imperative in nature. Major design flaws born of the original concessions contracts were altered during renegotiation, but only insufficiently so, largely because PAM Jaya and JWSRB lacked the authority, political power, and technical capacity to participate meaningfully. These flaws further corroded trust and limited opportunities for correction. The contracts have also failed to specify details regarding terms of measurement and modes of adjustment, precluding effective adaptation.

Jakarta also reminds us that the organizations involved in PPP are collections of actors with alliances, grudges, and rational and irrational ideas about other participants. Collectively, actors nurture institutional memories and strong ideas about “the other side.” The PPP began as a partnership forced upon an unwilling agency and challenged by inter-organizational cultural barriers. Despite numerous windows of opportunity to adjust, adaptability has been limited due self-reinforcing patterns of non-cooperation, poor information, and limited capacity for discretionary regulation – conditions impeding responsive regulation.

Currency, financial, and political risks (especially administrative changes) led to tariff freezes and other financial difficulties in both cases. The Manila concessions managed to adapt via contract mechanisms, such as EPAs, and the negotiation of extensions based on open reporting of impacts on earnings and agreed KPIs and BEMs. In Jakarta, however, targets were simply adjusted downwards to meet the guaranteed IRR with any kind of productive discussion of performance hindered by a gross lack of information, the sidelining of the regulator (who could have been a mediator), and corroded relationships amongst contract parties. Thus, the cases involve similar uncertainties but different outcomes, revealing the importance of responsive regulation, good information, and earned public legitimacy.

5 Conclusion: thoughts on managing uncertainty in water PPP

These cases offer several important lessons for dealing with uncertainty in water supply PPP, which apply beyond the cases to other water PPPs in developing regions globally. These lessons deal particularly with the role responsive regulation can play in mediating risks over the duration of a contract. First, decisions regarding the establishment of a regulatory body and the strength of enforcement have important temporal aspects. Reform follows a non-linear, iterative series of adjustments. In Manila, this is evident in the evolution of a regulatory body whose emergent professionalization and credibility have developed over time. This is particularly important in the water sector, where early information problems are likelier to demand adjustments once the PPP is operational. Early strong engagement legitimized the RO's role and allowed for more discretionary regulation later. Transparency, growing capacity, and public justification have allowed the organization to develop an effective model and assume an important role facilitating adjustment. The establishment of the JWSRB as an afterthought was too late, however: PAM Jaya and operators had established patterns of interaction that a weak RB could not supplant.

Second, contract ambiguity is an important tool for powerful and legally dexterous participants. Jakarta's frustrated efforts at rebasing and dispute resolution demonstrate the ability of stronger parties to gain control over contract ambiguities to out-maneuver weaker partners in disputes. This reaffirms the importance of contract specificity in the allocation of risks and rights, and particularly for decision rules when conflicts arise. Companies typically have more resources for legal support and can exert greater influence on dispute outcomes by pushing particular interpretations of ambiguous contract terms. In Jakarta, this was exacerbated by the absence of a neutral, legitimate third party to facilitate resolutions to contract incompleteness. In Manila, on the other hand, operators attempted to leverage ambiguities over the appropriate discount rate with less success, as they were better matched by a regulator with authority, good technical support, and wise use of windows of opportunity to clarify ambiguities.

Third, government effectiveness is critical to the public legitimation of adjustment, particularly for a more politically-sensitive public service, such as is water supply. In Manila, government capacity and increasing RO professionalization were stabilizing forces for the developing PPPs. Accommodating new information and dealing successfully with external shocks hinge on the capacity of the regulator to assess and audit operations and investments; identify problems; and justify decisions. In Jakarta, technical capacity problems limited PAM Jaya's ability to assess plans, and the RB's meet-in-the-middle approach was a weak substitute for meaningful deliberation. Jakarta also shows us that initial mistakes can be difficult to correct without sufficient information, public scrutiny, and capacity for technical assessment.

Fourth, capacity to manage uncertainty is undermined by weak relationships, distrust, and incoherence derived from cultural misunderstanding, secrecy, and intra-governmental conflict. This can be exacerbated by the symbolic importance of water. In Jakarta, cultural issues, legal bullying, and unproductive negotiations sustained patterns of non-cooperation. The experience suggests that adjustments are difficult when parties start at odds, hold different ideas about how water should be managed, or when the arbiter's capacity and power are limited. Because the agreements were insufficiently corrected, relationships devolved into unproductive stasis as players assumed defensive modes. Rather than renegotiating to a higher equilibrium, operators’ insistence on maintaining financially and politically unfeasible contracts shortened potential contract terms and limited expansion into adjacent markets. On the other hand, the government lost power to negotiate adjustments, as the probability of re-municipalization was tenuous due to lost government capacity to deliver water. Comparison also demonstrates the importance of transparency and public communication to making adjustments politically acceptable and dispelling concerns over corruption. In contrast to Manila, Jakarta's concessions, targets, and the details of regulatory decisions are almost entirely private. Lastly, the two cases reveal problems in intergovernmental conflict. Whereas Manila's agency was largely aligned with higher-level reform strategies, Jakarta's PAM Jaya was an unwilling participant whose efforts stood in clear opposition to the national government.

Finally, institutional “success” may be had without adaptability, but at the high cost of performance. The comparison of Manila with Jakarta is revelatory with respect to institutional versus instrumental performance. Maynilad's institutional ‘failure’ was, in the long run, Manila's success, whereas Jakarta's avoidance of termination has been an example of a PPP's failure to ‘fail’. Manila's unified national-local strategy, improving regulator, and consistent transparency all played a part in the salvation of the PSP experiment, where the converse was true in Jakarta. A prime concern in contract design should be insulating projects from political interference, but not at the expense of new information and direct responsiveness.

The cases demonstrate that responsive regulation depends on linkages between regulators, companies, and the public which give rise to a special kind of emergent knowledge, which is “localized, embedded, and invested in practice” (Weber & Khademian, 2008). This evolves through practice and is tied to identities, experiences, and relationships. Over time, collaborators can amalgamate knowledge and collectively generate new knowledge set specific to the problems at hand. This serves multiple purposes: creating better fit to context, uniting actors under common knowledge, aligning values and interests, and developing reasonable accountability systems. Lastly, consistency should necessarily not be a goal in and of itself, as its desirability depends on the quality of the agreement and how well the plan actually accomplishes its policy goals. Success, ultimately, must be judged according to instrumental performance and public value.

Notes

1 Risk may be defined as uncertainty with known probabilities associated with various outcomes. For the purpose of this paper, particularly considering that most outcomes related to PPP are difficult to assign probabilities to in reality, we take Irwin's definition of risk as “unpredictable variation in value” (2007) that can be used interchangeably with uncertainty. For more extensive discussion, see Irwin (2007) or Grimsey and Lewis (2004).

2 The Ayala group submitted bids significantly lower across the board, but bid rules precluded the same operators from winning both concessions.

3 The EPA would deal with changes in obligations, laws affecting cash flow, extraordinary currency fluctuations, or force majeure.

4 MWSS trustees would appoint regulators, and their decisions would be subject to approval of the President-appointed Board of Trustees.

5 At the time, regulators recall that they were “learning by doing” as they grappled with technical aspects of regulation and understanding their role.

6 The decision was made smoother by an interim downward revision of the requested increase.

7 Because capital expenditures were expected to be much lower for the established West, Maynilad had inherited 90% of the foreign-denominated MWSS debt. The justification was that the less populous and less developed East would require far more per capita investment.

8 Planned and actual expenditures deemed inefficient or imprudent would be disallowed from inclusion in the calculation of the companies’ opening cash positions.

9 The Appeals Panel upheld the RO's recommendations.

10 These included NRW, billed volume, daily production volume, collection efficiency, and customer service improvements.

11 Whereas the operators missed NRW by sizeable margins in the early years, they made great strides during the mid-2000s. Manila Water reduced NRW to just over 11% by 2012. Maynilad's NRW had risen until the second concession award, but the new consortium reversed the trend, bringing NRW down to 38.7% in 2013. While above Manila Water levels, the reduction was a notable accomplishment, given the short period.

12 Manila Water's 2012 rate was over nine times their bid in nominal pesos, having risen approximately 16.2% each year for fifteen years (MWSS Board Resolution No. 13-009-CA, 2013). Maynilad's 2012 rate was over six times the 7.21 bid and had risen 13.1% each year since 1997 (MWSS Board Resolution No. 13-01-CA, 2013. In hypothetical CPI-adjusted real 2012 pesos, the initial bid rates were approximately PHP8.86 for Manila Water and PHP15.90 for Maynilad. Even in real terms, this demonstrates a sharp increase of approximately 430% of the bid price for Manila Water and 286% for Maynilad.

13 Regarding the technical aspects, RO determined a PHP17.175 billion OCP for Manila Water, compared to their claimed PHP41.141 billion, and a PHP60.230 billion OCP for Maynilad compared to their claimed PHP67.213 billion (MWSS Board Resolution No. 2013-100-RO, 2013). The ROs also determined net present values of projected cash flows of PHP112.097 billion for Manila Water and PHP233.096 billion for Maynilad.

14 Perusahaan Daerah Pelayanan Air Minum Jakarta Raya.

15 For example, in 2010, PAM Jaya charged Palyja for missing a water quality target, but the contract specified no rules on frequency of monitoring. PAM Jaya suggested frequent oversight, whereas Palyja argued that enforcement be applied only annually (Interviews, JWSRB, 2013).

16 Though the RB technically continued to report to the Governor, they would now have to communicate via PAM Jaya (Interviews, JWSRB, 2013).

17 Despite warnings from the central government, PAM Jaya's leader, Maurits Napitupulu, continued to press a dispute over non-paying consumers. After delivering year-end notes that highlighted the potential debt of Rp 18.2 trillion due at the close of the contracts, the director was suddenly and unexpectedly removed from office.