Addressing the policy-implementation gaps in water services: the key role of meso-institutions

Gaps in the policy-formulation process

Genuine national control and monitoring of the policy process is essential for a successful policy development. In many cases, particularly in aid-dependent countries, governments suffer from pressure (either explicit or implicit) to develop reforms that might be neither truly demanded nor appropriately embedded in the social values of stakeholders. Another gap, partly linked to external pressure, is the development of overambitious policies, setting objectives that are disconnected from current national reality and capacities. While these policies might adhere to international thinking on good practice, the goals assigned result in unrealistic targets when it comes to their implementation, as well illustrated in the analysis of the Water Sector Development Plan 2006–2025 in Tanzania (Jiménez & Pérez-Foguet, Citation2011a). Basic aspects of capacity and institutional arrangements for delivery were hindering a sustainable service delivery in rural areas, which showed important weaknesses in terms of functionality (Jiménez & Pérez-Foguet, Citation2011b), quality and year-round availability of services (Jiménez & Pérez-Foguet, Citation2012). The National Sanitation Campaign, developed in Tanzania in 2012, shows similar weaknesses in the design and overambitious goals compared to current capacities (Jiménez, Mtango, & Cairncross, Citation2014a).

Another gap that might hinder policy implementation comes from lack of high-level political commitment (Savedoff & Spiller, Citation1999). In many contexts the volatility of political representatives within the government involves changes in priorities. Such volatility is particularly damaging in the water sector because of the amplitude of sunk costs and the long-term investments required, so that this sector faces a particularly challenging gap between the political and business cycles when it comes to the implementation of reform. Securing long-term political commitment is easier said than done, but it is essential in services such as water, which require long-term dedicated investments.

Building institutional devices that allow the inclusion of subnational governments, private and community stakeholders, making room for ‘voice’ in the policy-making process, is a necessary condition to craft realistic, implementable policies (Hirschman, Citation1970; OECD, Citation2011, Citation2015a). Particularly important in that respect is the capacity to tailor the participatory processes in a way that mitigates power imbalances between stakeholders and within communities (COHRE, AAAS, SDC & UN-Habitat, Citation2007), an inclusive strategy of particular significance when ethnic minorities and indigenous peoples are concerned (Jackson, Tan, Mooney, Hoverman, & White, Citation2012; Jiménez, Cortobius, & Kjellen, Citation2014b).

A partially related factor is the risk of policy capture by economic or political elites. Elite capture means that a few powerful and resourceful groups or individuals use their influence to usurp resources, skewing policy formulation or implementation (Shirley, Citation2002, Citation2008). Decentralization in the water sector has, for example, been criticized for failing to grasp properly social and political dynamics at the local level, which opens up the possibility of elite capture in local water users’ associations. For instance, there are many examples in developing countries where publicly funded bore-wells in rural areas are ending up in the backyard of a local strongman, preventing other villagers from making use of the well or obliging them to pay for rights of access (Rusca, Schwartz, Hadzovic, & Ahlers, Citation2015). Elite capture can also take place further upstream in the policy cycle. For example, water markets in Chile were captured by powerful buyers hoarding water rights that worked against more equitable water distributions to local farmers (Warner et al., Citation2009). Well-intentioned policy measures such as decentralization, public–private partnerships or market-based water allocations without the proper checks and balances and without adequate administrative capabilities (OECD, Citation2015a) may lead to local elites strengthening their positions at the expense of politically and economically marginalized groups.

Gaps in the operationalization of the policy

Once approved, policies need to be operationalized. Many obstacles can show up at this stage. In many occasions, the support of external agencies is focused on the policy-development phase, and the governments find themselves with approved policies that are not completely understood and with limited support, or even no support at all, for their implementation. In many cases, policy-making is ascribed to decision-makers, while implementation is linked to administrative capacity (UNESCO, Citation2006). One very common problem is the mismatch between the responsibilities allocated to certain ‘meso-institutions’ (local governments, regulatory agencies etc.), and the resources that flow into them (Crook, Citation2003; Jiménez & Pérez-Foguet, Citation2011c; OECD, Citation2015a; Ribot, Agrawal, & Larson, Citation2006). Beside the need for adequate funding, the human and technical capacity required cannot be acquired overnight. Policies establishing new roles for certain institutions or transferring new decision rights to devices in charge of implementing a policy need to deal with human resources that often takes time to train and adapt.

Moreover, the successful operationalization of policies depends on their perception as legitimate, which requires them to be clearly understood as well as effectively disseminated among all relevant stakeholders (OECD, Citation2015a; Stockholm Environment Institute (SEI), Citation2013). This requires a concerted effort to inform and train these stakeholders, particularly at the local level, which is often not planned for and might involve unexpected economic as well as political costs. The experience of users’ commissions in France (Commission Consultative des Services Publics Locaux) or in England and Wales (Consumer Council for Water) illustrates the difficulties to be faced, even in countries with relatively well-informed stakeholders.

A particular challenge impeding implementation is the misalignment between government-induced water policies and informal water institutions such as customary water rights. Failing to understand the role and functions of deeply embedded informal institutions and how they can be best included in the policy threatens its feasibility. For example, in the Central Asian former Soviet Union countries of Tajikistan, Kyrgyzstan and Uzbekistan it has been observed that there are competing institutional regimes based on recent integrated water resources management (IWRM) reform, an inherited Soviet managerial style of command and control, and customary water rights. Local actors typically combine newly established rules of local water management (water users’ associations) with informal institutions that often originate from institutionalized Soviet and pre‐Soviet patterns of behaviour (Schlüter & Herrfahrdt-Pahle, Citation2011; Sehring, Citation2009). This is a recipe made for confusion of which rules and regulations should apply, fuelling discretionary decisions in the operationalization of well-intentioned policies.

Path dependence and resulting misalignments can thus lead to formal institutions disrupted or completely replaced by informal ones, or to well-functioning informal institutions that are undermined. While some informal institutions can be well functioning and support reform aspirations, many are rather fuelling discretionary and clientelistic decision-making seriously distorting implementation. A case from Paraguay shows how better alignment between formal and informal institutions can benefit both sides. In Paraguay, informal private water-supply systems were recognized and agreements developed between local government entities and small-scale private water vendors. The outcome was easier control and monitoring of pricing and quality of service (Phumpiu & Gustafsson, Citation2009). Similar relatively successful combinations of formal and informal arrangements have been observed in other countries (e.g., Mozambique, Blanc, Citation2009; or Cambodia, Frenoux, Citation2016).

The underlying lesson is that successful policy implementation requires cooperative relationships between many parties operating within different institutional layers: government at various levels, the private sector and community-based organizations. When appropriate incentives to cooperate are not built into the policy framework, it is unlikely that actors that might have limited trust to each other will provide adequate support to policy implementation. Because of the centrality of water to human survival, suppliers, whether private or public, are often perceived as having disproportionate power, a perception often reinforced by information asymmetry among parties. Without institutions that have sufficient capacity to monitor and enforce agreed norms, and/or if users do not have adequate channels to signal their requests or express their dissatisfaction, incentives to implement welcome policies will likely be weakened or derailed, even paralyzed (OECD, Citation2015a).

Gaps related to the characteristics and behaviour of stakeholders

A part of these implementation problems is rooted at the micro-level, coming out of the variety of stakeholders involved in the provision as well as the usage of water, with actors that have different, often diverging or even conflicting, interests. Having appropriate institutional devices to monitor their rights and adequate institutional decision-making mechanism to coordinate their interests and eventually arbitrate among them are key factors in making a policy accepted or rejected.

On the supply side, an important characteristic of drinking water is the frequent monopolistic position of providers, at least in cities of a significant size,Footnote² which raises the classic problem of this market structure with respect to the delivery of adequate quantity at an appropriate price (OECD, Citation2011). The acceptance of institutional solutions to deal with these risks is a tributary of the type of regulatory arrangements and on the capacity of regulatory devices to benefit from adequate human and financial resources, to be able to constrain deviants and to resist capture by operators (OECD, Citation2015a).

However, the risk of capture is not solely on the supply side. Spiller (Citation2009) has shown that it can also come from what he calls ‘third party opportunism’, a situation in which interest groups that are particularly influential or which can build an active coalition may create powerful obstacles to the implementation of legitimate policies (e.g., preventing taxation of the benefactors of swimming pools in an environment with a scarce water resource) or distort policies through their influence on the decision-making process. The creation of representative organizations of consumers (e.g., water councils) does not automatically solve the problem since it depends so much on the quality of the representatives of stakeholders and the individual as well as institutional capacity to avoid capture by specific interest groups. This risk of third-party opportunism also raises the sensitive issue of possible conflicts between so-called ‘representative’ organizations and the responsibilities of elected decision-makers. Low participation of water users and the poor quality of those claiming to be their representatives, which is often fed by lack of resources and programmes to support their role, can therefore create gaps that they were intended to fill!

These different sources of opportunistic behaviour often end up in corruption, another and often overlooked source of the policy-implementation gap (e.g., Transparency International, Citation2008; Water Integrity Network, Citation2015). In many contexts corrupt behaviour has become the norm, fuelled by political patronage. Ironically, corruption can constitute an institution in its own right, following its own rules and regulations.

Corruption takes many forms, ranging from petty to grand or political corruption. Examples include:

• collusion between water consumers and local civil servants to rig water meters to avoid paying water fees, or consumers being pressurized to pay local civil servants for receiving the water service in the first place;

• bribery related to the awarding of licences for water uses and wastewater discharges;

• collusion (kickbacks or bid-rigging) and extortion in procurement procedures and the awarding of contracts;

• nepotism and kickbacks in the appointment and promotion to higher-level civil servant positions;

• collusion during the quality control of the construction of water infrastructure works; and

• embezzlement of government and foreign aid funds and assets

Endemic corruption undermines any water policy as well as administrative rules and regulations. It also erodes critical foundations of trust, the rule of law, fairness and efficiency of water institutions. Indeed, corruption and ‘red tapism’, often overlapping with bureaucratic inertia, increase transaction costs, discourage investments and feed strong disincentives for water-reform implementation. At its core corruption is a strong symptom of institutional weaknesses and poor governance. Recent studies confirm that even petty corruption at the water provider–consumer interface is one key risk that water sector must face (Butterworth & de la Harpe, Citation2009). This calls for more effective enforcement of regulatory institutions to monitor performance and expenditures of service providers (OECD, Citation2015a).

Gaps related to the overarching country-governance situation

Water policies do not exist in a vacuum; they depend on the overarching governance system (UNDP/SIWI, Citation2015). Several studies, going back at least to Shirley (Citation2002), suggest that changes in water governance are connected to other governance reforms.Footnote³

Indeed, water policies might be hindered by more general institutional factors that inhibit tentative reforms of governance (Jiménez, Jawara, LeDeunff, Naylor, & Scharp, Citation2017). For example, the establishment of regulatory body for water services in Bosnia and Herzegovina faced several obstacles derived from the complex administrative and political setup that prevented its implementation. Similarly, recent research shows how lack of accountability and poor top-down discipline within the public sector are among the major hindrances for effective service delivery in sub-Saharan Africa, including water (de Mariz, Ménard, & Abeillé, Citation2014; Overseas Development Institute (ODI), Citation2015). Fragile and post-conflict countries, which have among the lowest levels of coverage of water and sanitation services, face particular challenges. In most cases, the state has not the basic capacities to steer the sector, while the extension of service is heavily driven by external support, not always carried out in a very coordinated manner. High pressure to deliver on many different development priorities as well as the vulnerability to external shocks makes it even more complicated. For example, in Liberia, whose awful civil war ended in 2003, it was found in 2011 that over 95% of water points in the rural areas had been developed in the last decade by external agents, without coordination and control over conditions of delivery of water (Government of Liberia (GoL), Citation2011a), which in turns undermines the government’s capacity to ensure sustainability of the services. The governance of the sector is fragmented across various ministries, and government commitment to reform the sector, under the Liberia WASH Compact of 2011 (GoL, Citation2011b), remains to be implemented to a large extent.

The lesson from such cases is complex. On the one hand, delivering on policies to improve services can pay a double dividend in fragile states of enhancing well-being of citizens as well as gain state legitimacy (Brinkerhoff, Wetterberg, & Dunn, Citation2012). On the other hand, lack of adequate democratic culture, including debate, consultation and participation, represents a major challenge to the legitimacy of the policies adopted. More generally, improving governance in the water sector cannot be dissociated from the development of institutional devices through which stakeholders can be appropriately informed and can provide feedback to policy-makers. Building such institutions is a crucial element to effective reform, facilitating the implementation of policies chosen and reducing the economic and political transaction costs of that process.

Institutional layers: a much needed clarification

Indeed, much confusion remains when it comes to the specific components grouped under the terms ‘institutions’ and ‘governance’. The polymorphic meaning of ‘governance’ illustrates the difficulty: the term refers to very heterogeneous problems and serves various purposes (Tropp, Citation2007). For example, some contributions understand governance as the procedures needed to monitor rules of the game defined at the macro-level, such as the judicial procedures required to implement laws (World Development Report (WDR), Citation2017). In management, governance often refers to ‘corporate governance’, that is, modalities to pilot firms or other micro-institutions (NGOs, international organizations etc.). Still, other contributions interpret governance in the context of devices designed to implement policies understood as a set of rules, procedures and allocation mechanisms embedded in laws and regulation (OECD, Citation2015b; UNDP Water Governance Facility/UNICEF, Citation2015). These ambiguities are regrettable if one considers the importance of the issues as stake, which all relate to the implementation of rules. For example, recent OECD reports (e.g., OECD, Citation2009, Citation2015b, Citation2017) and numerous specific contributions to the WGI (e.g., the case studies on the Netherlands, OECD, Citation2014; or Brazil, OECD, Citation2015d), made a big step forward by taking on board the role of institutions in the development of policies and the explanation of their flaws. But they also suffer from these ambiguities, largely due to the mess in the representation of the different institutional layers in which ‘governance’ is embedded.

Building on Davis and North (Citation1971), Williamson (Citation1985), Aoki (Citation2001), Ostrom (Citation2005), Greif (Citation2006), among many others, recent developments suggest disentangling these institutional layers (Hodgson, Citation2015; Kunneke, Groenewegen, & Ménard, Citation2010; Ménard, Citation2017). In a sense, reports issued in the context of the WGI participate to this evolution through their effort to differentiate levels of government within which policies are elaborated and their governance implemented (OECD, Citation2011, Citation2015a, Citation2016). However, ‘levels of government’ is a limited unit of analysis when it comes to understanding policy gaps identified in the previous section. What matters most are the different institutional layers, of which ‘government levels’ are only a subset, within which these gaps are embedded.

Our starting point finds inspiration in the Northian perspective on institutions understood as ‘the set of fundamental political, social, and legal ground rules that establishes the basis for production, exchange, and distribution’ (Davis & North, Citation1971, p. 6) or, more succinctly, the ‘integrated systems of rules that structure social interactions’ (Hodgson, Citation2015, p. 57). ‘Institutions’ thus defined is therefore an overarching concept that captures different layers through which rules and norms are defined, implemented and operationalized.

In that respect, three layers can be distinguished. At the most general level are macro-institutions, within which constitutive rules and norms are defined and rights and duties established, delineating the domain of possible actions. For instance, the adoption of a law opening room for public–private partnership in water utilities typically belongs to this macro-institutional layer, while the entities created or restructured in that new legal context can be identified as micro-institutions (see below). Macro-institutions are ‘configurational’ (Ostrom, Citation2014): they determine who can hold what rights with what responsibilities, what can be done, and the constraints under which it can be done. In democratic regimes this is typically the role of the political system (e.g., a parliament) and/or the judiciary. Many policy gaps identified above in the second section come from flaws in these macro-institutions: poorly designed rules of the game such as badly defined property rights can introduce loopholes in procedures of implementation, opening room for opportunistic behaviour all the way up to bribery and corruption. Note that macro-institutions cannot be restricted to the sole formal, legalistic rules. North (Citation1990a) introduced an important distinction between formal and informal (macro-)institutions, a distinction relevant for the analysis of the water sector as emphasized in a UNDP report:

Social rules and norms matter because they shape beliefs and expectations that structure the behaviour of actors. Indeed, it is within these rules and norms that actors organize transactions needed for the production and delivery of water services, thus delineating the micro-institutional layer. This layer is the domain of water operators, water cooperatives, NGOs involved in the delivery of water and various users, actors that modern economic theory captures through the generic concept of ‘organizations’ (Gibbons & Roberts, Citation2013; Ménard, Citation2008; Milgrom & Roberts, Citation1992; Williamson, Citation1985, Citation1996).

Meso-institutions and the governance of water systems

However, there is a missing link in this picture, a gap between the layer at which general rules are defined and the layer at which actors make their decision in the context thus delineated. Indeed, macro-institutional rules and norms are rarely self-fulfilling and strictly implemented by micro-institutional actors. Rules and norms need to be ‘interpreted, ‘translated’ and ‘monitored’. For example, laws intending to guarantee more transparency in public procurement and to reinforce accountability might be adopted by a well-intentioned parliament; they still require devices to be implemented and monitored (de Mariz et al., Citation2014). Propositions regarding ‘The governance of water regulators’ (OECD, Citation2015b), or the need for a multilevel approach to governance in the water sector (OECD, Citation2011), typically relate to this issue. However, although such contributions rightly emphasize that there are different levels of government involved, they do not capture the role of devices often transversal to governmental levels and that play a central role in the implementation of policies. These devices fall into the category of what can be called intermediate or meso-institutions: this is the layer in which general rules and rights are interpreted, implemented, monitored and controlled. Regulatory agencies are such devices. Other examples are ministerial departments in charge of the water sector (typically a department of public works), local authorities responsible for the organization of the sector when water management is decentralized, self-organized local communities monitoring the resource etc. (Marques, Simões, Pires, & Almeida, Citation2009; OECD, Citation2015b).

Many policy gaps identified in the second section come from flaws in the design of these meso-institutions or in responsibilities they are allocated. Among the challenges they face is their capacity to operate both ways when bridging the gap between macro- and micro-institutions, facilitating the implementation of policies (top down) and channelling information about the social demand of users (bottom up). Another challenge concerns their capacity to adapt the general rules to specific situations in time, space and scope. Monitoring operator(s) in a city of several million inhabitants substantially differs from accomplishing the same task in a remote village, notwithstanding that both situations may fall under the same general laws. Similarly, general rules regarding tariffs may be misaligned with hydrological characteristics or users’ income of a specific region, thus doomed to fail in the absence of appropriate meso-institutions to adapt. Setbacks of many well-intentioned programmes of international organizations and development agencies are illustrative of failures to build intermediate devices that can adjust general rules to such context-dependent factors, as so well illustrated by the example of irrigation programmes in Nepal (Ostrom, Citation2014).

Most analytical approaches to meso-institutions have so far focused on regulatory agencies and the conditions of their efficiency, e.g., their independence from policy-makers, the availability of adequate financial resources etc. (Armstrong, Cowan, & Vickers, Citation1994; Laffont, Citation2005; Laffont & Tirole, Citation1993; OECD, Citation2015b; Spiller, Citation2009). On the empirical side, the diversity of institutional devices operating as meso-institutions is better acknowledged (OECD, Citation2011, Citation2015a, Citation2015b, Citation2016; Tremolet & Binder, Citation2010; UNDP, Citation2013). Efforts to develop a typology of these diverse arrangements can be found in Marques (Citation2010), the OECD report on water regulators (OECD, Citation2015b), and Ménard (Citation2017), among others. Table 2 summarizes (and illustrates) these institutional layers and their positioning at different administrative/geographical levels.

Building credibility to address policy gaps: political commitment and political transaction costs

The prevailing focus on regulatory agencies and the presumption of their superior efficiency tends to ignore factors that can make this solution fragile, threatening the implementation of water policies and the capacity to make them operational (see the second section).

First, one of the greatest challenges to efficient meso-institutional devices, particularly in developing as well as in ‘emerging’ economies, is the availability of competent human resources (OECD, Citation2015a, Principle No. 4). To be perceived as adequate and legitimate, staffing must be provided through clear standards with respect to: (1) entry-level qualification; (2) modalities of selection; (3) the appropriateness of remuneration; (4) the provision of a career path; and (5) the dissemination and enforcement of these standards (de Mariz et al., Citation2014). These conditions remain relevant whatever the type of meso-institutions at stake (regulatory agency, public bureau, local council etc.). Too many ‘light’ training programmes do not address this need to build human capital or, even worst, do not take into account the actual conditions under which staff operate, opening room to incompetent decisions … or corruption (see the second section; and OECD, Citation2015a, Principle Nos 4 and 9).

Second, water policies take time to be defined and implemented, which can undermine the credibility of policies and policy-makers. Indeed, a general policy problem, particularly acute in democratic regimes, is that political cycles do not match business cycles, making it hard for even well-intentioned policy-makers to adopt and stick to policies not matching their electoral cycle. In water systems, this lag between cycles is amplified by the long period required for the appropriate management of the resource and for recovering the sunk costs of the highly dedicated investments it needs (Savedoff & Spiller, Citation1999). The resulting difficulty of making commitments credible makes even more central the role of meso-institutions in the effective implementation of the policies adopted.

Third, building such meso-institutions involves significant transaction costs. Economic transaction costs are the costs of planning, monitoring and bringing to its achievement the allocation and transfer of rights to use the resource or its associated services (Williamson, Citation1996). Examples are the costs of contracting with a private operator, or monitoring the delegation of management to local authorities. Most analyses of these costs have so far focused on the micro-institutional level, e.g., the costs and benefits of operating through contracts between a public authority and a private operator rather than offering the service through a public monopoly.Footnote⁴ However, there are also costs attached to meso-institutions, such as those of running a regulatory agency, the bureaucratic costs of a public department monitoring water services etc. These costs may well challenge the appropriateness of the solution selected to implement water policies (for illustrations, see Ostrom, Citation2005, Citation2014). Unfortunately, there are very few such comparative assessments so far (but see Libecap, Citation2014, and Vatn, Citation2015, for environmental policies; and Ménard, Citation2016, for fisheries).

Fourth, many policy gaps identified above are coming out of what North (Citation1990b) called political transaction costs (see also Marshall & Weingast, Citation1988; Robinson, Citation2010). With respect to meso-institutions, political transaction costs are the costs of building and monitoring coalitions among actors in charge of implementing the rules and norms that organize the sector. For instance, once a legal framework has been adopted that favours a public–private partnerships to develop or modernize water infrastructures, there are costs to build and implement an agreement between, say, local authorities, a private operator and representative of users (e.g., consumers’ councils). These costs might create gaps between what is expected from the policy adopted and the conditions of its implementation. They might even be prohibitive, as when the behaviour of some parties creates rigidities that disrupt the development of a project, what Spiller (Citation2009) identified as ‘third party opportunism’ and risk of capture by specific groups of interests (see above).

Unfortunately, there are so far very few empirical studies assessing these costs associated with the existence and running of meso-institutions (but see Grafton, Libecap, Edwards, O’Brien, & Landry, Citation2012). However, numerous indicators have recently been proposed in the water sector to identify these costs and measure them (see the pioneering contribution by Araral & Yu, Citation2013; Ménard, Citation2016; OECD, Citation2015c, Citation2016). Developing these tools should be high on the agenda of researchers and organizations working on policy gaps in water services (as well as in other ‘essential services’).