A Political Economy of Privatization Contracts: The Case of Water and Sanitation in Ghana and Argentina

In general, the process and outcomes of privatization have been studied from the point of view of efficiency. In this article, we consider issues in the course of contract design, implementation, management and enforcement in privatized public services and utilities. The study is based on two case studies, involving several water concessions in Argentina and a management contract in the urban water sector in Ghana. Three key arguments are presented on the basis of these case studies. The first is that an individualistic analytical framework is often utilized by the mainstream economic perspectives, but these are inadequate for a comparative assessment of private versus public provision in public services where there are distinct collective or group interests and hence a wider socio-economic context and representation of different interests becomes highly important. Instead, the article proposes a political economy perspective, which pays due attention to distributional issues, group interests, ideology of states and power relations for the assessment of privatization contracts. Second, the administrative capacity of states and their resources play a key role for the outcomes of privatization. Finally, while some contractual issues could be resolved through resourcing and experience over time, others are inherent to the contractual relations with little prospect of remedy.


Introduction
Since the early 1980s the privatisation of economic activities across the world has taken place in three phases. The first started with the privatisation of manufacturing, banking as well as agricultural activities (especially in the developing world) and was more or less completed by the end of 1990s.
This was followed by the privatisation of infrastructure and public utility services, starting around the early 1990s and continuing to the end of the 2000s with some set back after the 1997 financial crisis in Asia. The third wave of privatisation in different forms is now on-going for core public services such health and education. This process has been accompanied by the emergence of a prolific literature on the outcomes of privatisation. Most of these studies have assessed the process on the basis of various performance indicators such as profitability and efficiency, involving output and cost comparisons (Megginson et al., 1994;Megginson and Netter, 2001;Goldstein, 1999, Villalonga, 2000Florio and Manzoni, 2004;Parker and Kirkpatrick, 2005).
In this article, we focus on the privatisation of public services and infrastructure and assess the process, using service contracts as unit of analysis. The private sector is often brought into the utility and public services such as health and education, water and electricity through long-term contracts.
Private finance contracts to build and operate hospitals and schools, concession, lease and management contracts to supply water and electricity have been used across the world. 1 However, contractual failures involving renegotiations, protracted disputes and cancellations have been widespread in many countries (Guasch 2004;Dagdeviren and Robertson, 2011). Examining the process and outcomes of privatisation of services through a focus on the design, implementation and outcomes of service contracts can often provide invaluable insights that cannot be acquired from the conventional studies of performance and efficiency assessment. Studies on service contracts, for example, can reveal crucial evidence on opportunistic behaviour, the nature of 3 conflicts of interests and the institutional capacity in the design, implementation and enforcement of contracts, all of which have crucial bearing for service performance and users' welfare.

The discussion and analysis in this article is based on primary research in Argentina and in
Ghana. The first is a case study of four concession contracts in water and sanitation in Argentina, the second is a study of management contract in the water sector in Ghana. Although there are substantial differences between these countries in terms of their socio-economic, political and institutional development considerable commonalities exist with regard to the contractual problems they face in privatised public services. The study highlights the problems encountered in the design, application and enforcement of contracts and suggests that while some of the issues could be managed and improved upon with experience, extra resources, better monitoring and enforcement, there are inherent problems in contractual process for which little or no remedy exists. However, beyond contractual problems the analysis is broadly informed by a political -economy perspective by rejecting the methodological individualism that underpins mainstream economic perspectives. In addition, the role of the state is emphasised in being subject to pressure from different interest groups and also swayed by an ideological agenda not substantiated by evidence on privatisation, rather than as an impartial, arbiter of interests.

Mainstream theoretical approaches to the privatisation of public services
Broadly speaking, mainstream economic theory has been against government ownership and control of resources on grounds of efficiency considerations. The exceptions, as are often cited in textbooks, are cases where there are market failures, for example if the sector is a natural monopoly, or it generates externalities or produces public goods. As market failures characterise many of the public services and utilities, the position of the mainstream theory has therefore been in favour of public ownership for these sectors. However, there have been some detractors from this position, one of which is Public Choice Theory (PCT). Applying methodological individualism to the 4 political and public sphere (that is assuming rational, self-interested bureaucrats), PCT counterbalances the emphasis on market failures with a focus on 'government failures' (Buchanan, 1978), arising from lack of incentives for cost cutting (Niskanen, 1968 and1976), problems associated with majority voting in public decision making (Tullock, 1959 and1962) and rent-seeking (Buchanan, 1983).
Secondly, property rights theory (PRT) posits that privatising economic resources in general would resolve the inefficiencies associated with market failures. It was suggested for example that the externalities arising from collective ownership, such as the overutilization of resources; freeriding on investment and innovation, difficulties of measuring, attributing and rewarding effort correctly, could be internalised through privatisation (Demsetz, 1967;Demsetz, 1972 and1973;Furubotn and Pejovich, 1972). Competitive bidding and rivalry (Demsetz 1968 and1971) and independent regulation (Goldberg, 1974;Shirley, 2002) or use of anti-trust laws (Posner, 1974 and are recommended for addressing market failures such as those arising from monopolistic market structure as in utilities. Further extensions to the PRT have been made by incomplete contracts theory (Grossman and Hart, 1986;Moore 1988 and on the grounds that the former does not provide sufficient guidance on the forms of ownership when contracts are incomplete. This theory predicts a superior performance under private ownership for services such as water and sanitation, health, waste collection and prison services. It is proposed that some efficiency improvements in the form of cost reductions and/or quality enhancement are non-contractible ex-ante because costs and benefits are non-verifiable (though observable). It is assumed that any non-contracted improvement by public managers is subject to approval by bureaucrats and the results will not fully translate into direct benefits. This is called 'the hold-up' problem in the public sector where incentives for innovation are considered to be lower. Privatisation is expected to resolve the hold-up problem on the basis of the assumption that activities leading to efficiency gains would not be subject to approval in the private sector. On the basis of this analysis, the proponents of the incomplete 5 contracts theory (TIC) reach the general conclusion that private arrangements increase efficiency unless the social gains from cost reduction is small and they are offset by the negative impact of cost cutting efforts on quality (Hart et al 1997).
While the theory of incomplete contracts recommends privatisation of public services to deal with the 'hold-up' problem in the public sector, the evidence in this paper show that privatisation replaces one incomplete contract (between state and public manager) with another one (between state and private provider) and the possibility of hold-up by any of the parties is not eliminated. The incessant disputes and renegotiations over contractual terms and conditions (e.g. tariffs and investment in Argentina, NRW in Ghana) can quite often be initiated by the private sector and these can be seen as a form of hold-up.
Another relevant theoretical framework for privatisation is the transaction cost theory (TCT) which is built on the Coasian (1937) critique of neoclassical theory. There is wide recognition in this literature that property rights, contractual hazards, uncertainty and information imperfections (incompleteness and asymmetry) matter and give rise to transaction costs (Williamson, 1971, 1979, 1981, North 1990. The basic premise of the theory is that trade through the market is best when transaction costs are low. Otherwise, integration within a firm helps in reducing the cost of transacting through markets. While the scholars within TCT tradition have been generally in favour of 'private order', the principal elements of TCT do not provide an unambiguous rationale for privatisation, particularly for public services. This is especially because the factors that give rise to higher transaction cost (for example incomplete information, contractual hazards and uncertainty) are not unique to the public sector and do not disappear with privatisation and in fact in some cases transaction costs increase under the private model.
Finally, the fact that business activities are usually carried out in firms where ownership and management is separated leads to the recognition of agency costs associated with delegation (Fama and Jensen, 1983). These arise from asymmetric information between owners and managers with different objectives and give rise to what is known as 'principal-agent problem'. The principal-agent 6 theory does not make a uniform suggestion in favour of public or private ownership for minimising delegation costs. Instead, its guidance is that ownership decisions should be informed by incentive structure that depends on the information asymmetries and agency relations (Sappington and Stiglitz, 1987;Shapiro and Willig, 1990;Laffont and Tirole, 1990). Despite the absence of a strong theoretical rationale for privatisation of utilities and public services, an unfettered tendency in that direction on a global scale has gradually been taking place.
In the next section, the discussion around contractual implementation in two case studies from a low and middle income country aims to show failures of privatisation in public services. This enables a discussion of the shortcomings of mainstream economic theory in understanding contractual problems and points to the necessity of a wider political economy approach.

Argentina: Post-privatisation contractual disputes in the water and sanitation sector 2
Argentina was one of the first countries that begun privatising its utilities extensively from the early 1990s, largely through foreign direct investment with the involvement of Multinational Companies (MNCs), as part of a wider set of neo-liberal reforms that started in the 1980s. The country had entered into Bilateral Investment Treaties (BITs) with more than fifty other states and the guarantees given to MNCs by these agreements played an instrumental role in attracting foreign investors. A severe economic crisis in 2001 and the abandonment of pegged exchange rate affected the profitability of the companies and started a process of disputes and renegotiations in which over sixty MNCs sought protection under the BITs.

7
In the water and sanitation sector six concession contracts with thirty years of duration were cancelled. In this case study, we discuss four of these: The analysis draws on the information contained in sixteen claim and counter claim files submitted by the private investors and the government of Argentina to an international arbitrator. 3 Although the narrative in these files is subjective, reflecting the interests of individual parties, in some cases the claims of contracting parties overlap, which aids verification of the facts.

Common features in the narrative of dispute files
Firstly, issues of tariffs, investment targets and strategy of financing investment were the main subjects of disputes in all cases, reflecting a contest over the allocation of risks and profits. In some cases, the dispute was related to the differences in the interpretation of contracts. For example in one case, in contrast to the public agencies' interpretation, the operator insisted that the contract allowed for taxes to be passed on users through prices. In other cases, contracts permitted various forms of increase in service charges which in practice caused social unrest and became politically unsustainable. Hence, companies asked for a reduction in investment targets in return for a more socially applicable tariff structure. Such social (and distributional) conflicts have been widespread especially in the developing world but these issues are not tackled by the theories covered in previous section because they focus on firm level efficiency.
8 Secondly, in cases where disputes arose following the economic crisis, a major disagreement was related to the accumulated external debt, which the companies had used to fund their investment with very little equity in the 1990s when interest rates were low and exchange rate was fixed. After the crisis, a substantial devaluation threatened the financial viability of these companies as they had to pay their debt in foreign currencies, while earning their revenues in domestic currency. Hence, companies requested adjustment to their tariffs in order to service their debt which the government refused, pointing to the fact that investment funding strategy (i.e. whether funded by external or internal resources) was a business decision and the associated risks should be assumed by the companies that took such decisions.
Finally, one of the common features of the narratives in the dispute files was the prevalence of opportunistic behaviour by contractual parties from the beginning to the end when the contracts were terminated. For example, public authorities disregarded user opinion in the process of privatisation in that plans for large scale privatisation of utilities were implemented as a matter of political decision without consultation with user and civil society groups. However, after the emergence of disputes, they exploited public opinion when necessary, for example during the arbitration proceedings. Similarly, the request by some companies to renegotiate the terms and conditions of the contract after winning the concessions by submitting the most favourable bids (that is that those with lowest tariffs or highest concession fee) could be seen in the same light.

Root Causes of the Disputes
Several factors could be highlighted as the fundamental causes of the disputes involving concession contracts in water and sanitation services. First of all, the four cases of disputes provide compelling evidence that these essentially boil down to social conflicts of interest over the supply of a vital service (such as water and sanitation). Social unrest and political campaigns to reverse privatisation and penalise private companies because of high service charges and a number of 9 incidents when water supply was affected by turbidity reflect the response of user groups when their interests in affordable and good quality water are damaged. Bakker (2010, p.140)  It is usually presumed that public agencies represent the interests of users (as well as others), but this is highly contentious when governments are committed to privatisation in such a way that they concede to the demands of operators in the process of renegotiation in order for privatisation not to fail. The problematic nature of representation by the governments becomes clear when the predominant ideology and predetermined mandate with respect to privatisation is considered together with; the involvement of the powerful MNCs; the absence of user participation in the process of formulating contractual obligations and targets; and the fact that contracts are often seen as commercially sensitive documents with great secrecy and no transparency about their details.

10
The social unrest caused by the increases in service charges (for example those related to additional infrastructure charges) that later proved politically inapplicable reveal either a lack of understanding or negligence, on the part of public authorities, of issues related to the affordability of services charges and expectations of users since some of these increases were permitted by the contracts. Later, in their counter claims submitted to the arbitrator against the claims of the disputing companies, public authorities utilised a legal argument about the 'defence of public necessity', reflecting the afterthoughts on matters such as affordability and social context of service provision.
Finally, defects in the design of contracts were widespread. In some cases contracts failed to specify the consequences of non-compliance, for example, with investment targets. In other cases,

Ghana: Management Contract in the Water Sector 4
Ghana contracted out its water supply services for five years from 2006 following a major restructuring in the sector that aimed to improve its attraction for the private investors. For this purpose, the rural water supply and sanitation services (the unprofitable elements) were separated and retained within the public sector. The contract was awarded to a joint venture company of the Dutch Aqua Vitens and South African Rand Water. The privatisation process had been supported by various donors, especially the World Bank with grants and technical assistance. The Management 11 Contract described the obligations of the private operator in some detail the most important of which were the following: a) A minimum five per cent reduction per year in non-revenue water (NRW), i.e. leakages, nonpayments, illegal use etc. b) Increasing revenue collection and reduction in chemical and power consumption c) Maintaining water quality standard and improve customer response d) Lowering public sector's water consumption Throughout the contract period, these targets were not achieved and obligations were not fulfilled by the private operator, except for some reduction regarding the consumption of electricity and faster response to customers' calls. Although there was some recovery in revenue levels this was due to considerable increase in water charges rather than better operational performance.
Further, there was extensive non-compliance with water quality requirements. As a result, when the contract period ended, the government took the water supply service back into public ownership despite the initial intentions of longer term privatisation.
While the experience with and outcome of privatisation could be assessed from different perspectives, the contractual underpinnings of poor performance lay in several areas. The foremost defect of the contract was that it had not determined the baselines against which the performance of the operator could be assessed. Instead, it had left these baselines to be determined in consultation with the operator after the contract was awarded. For example, although the contract required the management company to reduce NRW by a minimum of five per cent every year, from what level this reduction were to be achieved had not been specified. No agreement could be reached on most baselines in the five years when the management contract was in force.
Renegotiations between the public counterparty and the private company proved futile since the private operator was better off with a defective contract (which made the assessment of its performance impossible) and guaranteed management fee. Moreover, the contract left the determination of penalties for the breach of contractual targets and requirements to private operator. As would be expected, these penalties were not specified and hence could not be enforced despite widespread non-compliance.
Secondly, as in the case of contractual issues in Argentina, the water management contract in Ghana also contained many clauses open to interpretation. For example, the contract obliged the operator to make certain decisions 'in consultation' with the public counterpart. However, it was not clear how the ultimate decision would be made when consultation did not result in some form of negotiated agreement. Similarly, the public counterpart was defined as the 'supervisor' of the operator in one part of the contract, while the operator had been defined as 'independent' in another part. These inconsistencies created considerable conflict and affected the operations in a negative way. Moreover, in determining the payment priority (for salaries, inputs, repairs and maintenance, transfer of surplus to the public counterpart) the contract required the operator to be 'prudent' in spending and cover 'reasonable and necessary costs'. However, the vagueness of these terms resulted in spending patterns that caused significant hostility on the part of the public authority due to lack of any surplus transfer for prolonged periods of time.
Finally, the companies bidding for the operation of water services had submitted detailed plans about finance and personnel, including the list of senior managers and their resumes. However, the public counterparty argued that the management company had not brought in the skilled personnel promised during the bidding process. The company could not be held accountable for this failure as the contract indicated that the only binding document was the contract itself and that all prior documents (bidding documents) were null and void.
Overall, the privatisation experience in the Ghanian urban water sector was characterised by capacity weaknesses in the design, application and enforcement of the management contract.

Discussion
At a more abstract level, various elements of the case studies discussed in this paper reveal the relevance and importance of a political economy perspective in understanding the consequences of 13 privatisation of public services like water and sanitation. Of the numerous political economy dimensions that could be highlighted, two are most imperative for the assessment of privatisation programmes: a) treatment of social and distributional conflicts, b) the ideology and capacity of the state. Let us consider these in turn.

Treatment of social and distributional conflicts
Most of the theoretical arguments discussed in section 2 are based on what is known as 'methodological individualism' in that they focus on transactions between self-interested individuals.
Wider socio-economic and political context and the distributional consequences of economic policies are ignored. Social or group interests which prevail for public services are not considered in these theoretical formulations. The case studies in this paper show the crucial importance of such interest and limitations of frameworks that utilise individual-based analysis. For example, in services where user charges apply as in the case of water sector in Argentina and Ghana, increases in tariffs to ensure cost recovery and profitability affected the welfare of users, especially the low income groups and the poor negatively. An important implication of the approach in mainstream theories is that they are capable of dealing with conflicts between individual parties to a transaction but they are not able to deal with social conflicts arising from such transactions. Social conflicts over service supply or charges inevitably lead to politicisation of service delivery in countries with a relatively democratic tradition, and reversal of private model is likely if social and distributional context is not carefully considered as in Ghana and Argentina where social discontent with the outcomes of privatisation led to termination or non-renewal of the private contracts.
A particularly challenging matter, it seems, is related to how the conflict of interests between users and private investors (or between service quality, access, affordability and cost recovery and profitability) are addressed in the contracts outsourcing public services to the private sector. The complete lack of transparency about contract design, development and accessibility does not reflect 14 well on the process in terms of user participation and accountability to the public. It is almost a universal international practice that the public service contracts are designed and applied often without user involvement and they are not available for public scrutiny despite varying levels of broader oversight by consumer groups and regulators after privatisation. The fact that the so called 'commercial sensitivity of contracts' takes precedence to public interest in essential services reflects a bias in favour of private investors. Although it is presumed that the state and its agencies can sufficiently and comprehensively represent the interests of a wide range of stakeholders in the process, this is rather dubious as reflected by the case of Argentina where tariff structures and levels agreed by the public authorities and permitted by the contracts proved to be socially unsustainable, or Ghana where allegations of corruption derailed the earlier privatisation plans.
Moreover, contractual disputes significantly increase the transaction costs associated with privatisation. While the knowledge about these costs is important for the assessment of the wider social benefits of privatisation programmes this information is not often in the public domain unless the disputes are channelled through formal institutions (i.e. courts or arbitrators). Four of the dispute cases in Argentina, for example, took years of legal proceedings and all of these disputes are still pending. The claims of the companies are over 1.5 billion USD which if successful has to be added over and above the initial transaction costs of privatisation. In cases where disputes are resolved through renegotiations compromise from initial targets is common. Contractual disputes and renegotiations are very likely in a number of circumstances. External shocks increase the potential for conflicts if they affect contracting parties' costs and benefits. Lack of due diligence in the assessment of the contractual rewards and liabilities can lead to disagreements in the aftermath of privatisation. Ineffective representation of users in determining the contractual parameters that affect their welfare can also fuel social opposition.
The ideology and the capacity of the state From a theoretical perspective, the pro-privatisation arguments have been based on the assumed superiority of the private sector in terms of efficiency. At a practical level, investment in public services have also played an important role for privatisation of public services and utilities in an environment in which the arguments for a reduced role for the state have gained a stronghold position, leading governments to privatise cash strapped services and relinquish their responsibility for looming investment requirements.
However, both the broader evidence and the evidence in this paper indicate that the investment by the private sector in utilities and public services has not always been at levels expected and smoothly forthcoming. The states have continued to invest a larger proportion of total capital in infrastructure and public services. In the developing world, for example, it was shown by Estache (2006)  This argument is also supported by Guasch and Straub (2006). The evidence from the case studies above reveals the problematic nature of privatisation for investment in public services. The case study on Argentina shows that the private water companies have reneged on their investment targets relied excessively on debt for funding investments threatened the financial viability of operations and resulted in state either taking over the services or injecting public resources to ensure their viability. The increasing tendency in African economies for the use of management contracts that retain the risk of investment in the public sector unlike concessions and PFIs, reflect the deviation between the rhetoric and reality in dealing with investment needs in public services.
The case studies of Argentina and Ghana provide useful insights into the nature of states' agency. Most significantly, they indicate that the states, its governments and institutions are often subject to deeply entrenched ideological influences that displace coherent policy choices. The ideological commitment of the governments to privatisation in general meant that the service users have not been consulted and opposition to these programmes have been disregarded in Argentina and Ghana. Lessons from the widespread failures in the privatisation of public services in other countries have been ignored.
Neither are the states and institutions neutral mediators of divergent interests held in the provision of public services. The political domain in which they operate is subject to contestation and pressure from different interest groups including powerful political lobbies and civil society organisations. However, they are not passive respondents to such interactions. The transformation towards more neo-liberal and capital oriented political regimes in the last several decades led governments to show an ideological bias in upholding the interests of private capital. This is partly why the commercial sensitivity of public service contracts supersedes the public interest in their disclosure. This is also why the regulators in Argentina felt that they have to 'give in' to the demands of private investors in order not to tarnish the image of privatisation.
The case studies also highlight the importance of administrative capacity of state in terms of the outcomes of privatisation of public services but this is ignored by some of the theories covered above. The evidence on the complex and troublesome nature of contract design, application, management and enforcement before and after the privatisation of public services and utilities is compelling. These include the difficulties in ensuring consistent and unambiguous contractual clauses on rewards and liabilities (e.g. investment vs tariffs in Argentina), roles and responsibilities of contractual parties (e.g. in Ghana) that are suitable for current as well as future periods with possible changes in the operational environment.
Performance of services under private contracting depends on the presence of effective monitoring and enforcement systems. These include systems that guarantee the availability of necessary information for performance assessment through appropriate reporting forms with adequate frequency and good standards; institutions and capacity for effective oversight over contractual obligations and enforcement mechanisms and procedures for contractual noncompliance. The study of Ghana demonstrated how acute information problems due to absence of effective reporting systems made it impossible to apply any penalties to the water company. In both countries there is evidence of ineffective enforcement in relation to contractual noncompliance. For example, in Argentina investment targets were not fulfilled by water companies, so was NRW target in Ghana. Lack of availability of service contracts for public scrutiny also weakens the potential for monitoring such contracts.
It is clear that contracts that outsource public services to private sector can be better designed and implemented with greater experience, resources (skills, financial) and effective monitoring and enforcement systems . Nevertheless, there are a number of inherent problems in the design and management of contracts with little or no remedy. Firstly, it is impossible to write fully consistent, unambiguous contractual clauses that provide optimal incentives to all parties. Secondly, and as is widely acknowledged, long-term contracts are inflexible and uncertain. What transpires in the future can change the projected cost and benefits of contracted services for different stakeholders, lead to suboptimal outcomes and create winners and losers. This becomes particularly obvious in times of crises and following the downturn in the investment cycles. Thirdly, in privately provided public services the clash of private with public interests is ubiquitous. Fourthly, opportunistic behaviour by contractual parties can be reduced but not eliminated. For example, if disputes and renegotiations generate higher rents by reneging on responsibilities, they are likely to prevail. Finally, there is no permanent 'risk transfer' in outsourced public services given that governments have to pick up the pieces for failed public service projects as in Argentina and Ghana.

Conclusions
In this paper, privatisation of public services has been discussed on the basis of contractual issues encountered in the process of implementation. The conclusions are derived from two case studies from water sector in Argentina and Ghana. Four major points from the foregoing discussion deserve emphasis. First of all, social and distributional outcomes of provision are in no other sector play as decisive a role as they do in public services. Any adverse change in service delivery, affecting the access of households, following privatisation has to be set against firm level efficiency gains. In certain socio-economic contexts such distributional conflicts can deal a serious blow to the viability of privatisation programmes.
Secondly, representation of different interests in the design and administration of public service contracts seems to be rather problematic. While private investors do have direct involvement and influence in the process, the interests of users, especially the low income and the poor, are presumed to be represented indirectly by the agencies of the state. However, effectiveness of this representation is rather unconvincing in a world in which the selection of economic policies is significantly influenced by a political ideology aiming to enhance the penetration of private capital into areas that had been traditionally dominated by the public sector. The case studies above provide explicit or implicit examples which raise questions about the impartiality of public agencies' representation.
Thirdly, the capacity of states seems to be a crucial factor in the design, implementation, monitoring and enforcement of public service contracts. Both in Argentina and Ghana capacity weaknesses arising from institutional problems or resource shortages resulted in flawed contracts and weak enforcement.
Finally, while some of the issues raised in this paper could be dealt with targeted interventions (greater resources, experience gained over time), others are innate to privatisation and cannot be eliminated fully and permanently. For example, it is impossible to write fully consistent, unambiguous contractual clauses that provide optimal incentives to all parties into the foreseeable future.