* SPRU - Science and Technology Policy Research, Mantell Building, University of Sussex, Brighton BN1 9RF, UK. Email: A.G.Smith@sussex.ac.uk.

Published in Journal of Environmental Policy and Planning, 2000, Vol. 2, No. 2, pp.115-134.

Abstract

This paper analyses implementation of the 1991 Urban Waste Water Treatment Directive (UWWTD) in England and Wales. In its Fifth Environmental Action Programme (EAP, 1993-2000), the European Commission committed itself to improving the implementation of EU environmental policies. It has been suggested that implementation will be more effective when the fit between Directive imperative and Member State administrative tradition is close. Unfortunately, the fit approach sheds little light on the processes by which an accommodation is reached between Directive and Member State. The approach to UWWTD implementation taken in England and Wales has recently shifted, changing from a minimalist approach to a more precautionary approach. This change affords a useful opportunity to analyse the processes by which Member States and Directive requirements mutually adjust during implementation. In this case, these processes are explained by broadening the analytical sweep to include the machinations of relevant Member State policy networks and the influence of outsider policy beliefs that orbit those networks.

Introduction

In its Fifth Environmental Action Programme (5th EAP, 1993-2000), the European Commission committed itself to improving the implementation of EU environmental policies (O.J. C138 17/05/93). If EU policy is to be more than a paper exercise, then tangible results on the ground must be the final arbiter of those policies (Jordan, 1999, p.72). The Commission realise that policy success rests as much upon effective implementation beyond legal transposition as it does in the initial design of policy instruments. What influences the national implementation of European Directives? And how do these processes shape final Directive outputs? This paper seeks to answer these two questions for the case of the 1991 Urban Waste Water Treatment Directive (UWWTD; CEC, 1991) implemented in England and Wales. In so doing, it explores in greater detail the notion of fit that has been developed by others to explain implementation of European Directives (Knill, 1998). Essentially, the better the fit between Directive requirement and national administration, the more effective the implementation (i.e. it is the fit that influences Directive outputs). This view implies 5th EAP ambitions will be met if the Commission and Member States improve the fit between Directive and national administrations.

The primary purpose of this paper is to present an empirical study. But this study is presented in such a way as to problematise narrow approaches to fit, and to explore the case from a broader perspective. This view implies practical Directive requirements are not set out in detail a priori, and that they are open to a degree of bargaining at national and sub-national levels.

The UWWTD was transposed for implementation in England and Wales in 1994, so its implementation has been contemporary to the 5th EAP. The case study is interesting because the approach taken in England and Wales changed recently: from a minimalist interpretation to a more precautionary approach. This change affords a useful opportunity to analyse influences upon Directive implementation. Moreover, the UWWTD was pitched into a highly unstable English and Welsh regulatory system, in which administrative traditions were in a state of flux following privatisation of the water industry in 1989. This makes the Member State a suitable choice for studying processes whereby the implementation of Directive imperatives and national circumstances co-adapt. Earlier studies have explained why the UK yielded to Europe and agreed to the UWWTD (Jordan and Greenaway, 1998; Ward, 1998), but none have really analysed how it has been implemented. What is remarkable about the English and Welsh case study is that its implementation is inexplicable without analysing the economic regulatory framework. The influence of this non-environmental regulatory regime in this case is illuminating for Member States currently liberalising their own utility services.

The fit between Directive imperative and national administrative tradition

Fit compares Directive requirement with existing administrative systems in Member States. Fit, so the argument goes, explains the difficulty of adaptation between national administrations and Directives and hence the effectiveness of implementation. It hypothesises that the better the fit, the more effective the implementation, i.e. the more readily Member States adopt and act upon Directive requirement (Knill, 1998, p.2). Christoph Knill has developed this notion and measures fit by comparing Member State administrative traditions against the nature of requirements in the Directive (ibid, p.4). Accordingly, it is these two variables which will answer our research questions concerning national influences upon Directive outputs (i.e. national administrative tradition and Directive requirement). The greater the mismatch, the higher the degree of administrative adaptation required by Member States (ibid, p.2) or, conversely, the stronger the effect of national institutions on the implementation of the Directive.

Unfortunately, as Knill rightly concludes, the fit model cannot tell us a great deal beyond anticipating effective implementation when there is little adaptive pressure, and less effective implementation in cases of high adaptive pressure (1998, p.25). Ironically, if there is a good fit then there may have been little need for the Directive in a particular Member State in the first place. The Netherlands was already meeting the requirements of the UWWTD when it was being negotiated, for a variety of domestic reasons (Kelder, 1999). Thus Directives need sometimes to be viewed as a device for moving laggard countries closer to the vanguard. With laggard Member States we can expect at least a moderate degree of misfit; and it is the processes by which the national administration changes to fit Directive requirement that is interesting analytically.

In this respect, the fit approach implies a top-down view of implementation, i.e. the process of implementation will involve authoritative Member State policy-makers at the top forging a hierarchical ‘implementation chain’ of subordinate administrative units to deliver Directive requirements on the ground (Pressman and Wildavsky, 1973). Essentially, the challenge to Member States is to provide the necessary administrative units and personnel (e.g. the environmental regulatory body and its inspectors) with sufficient resources and clear instructions that will deliver given and accepted Directive requirements (Mazmanian and Sabatier, 1989, p.290). Implementation is assumed to be a fundamentally managerial task. This is a view held by some policy practitioners (Rhodes and Marsh, 1992), and we can see evidence of this in the way the Department of the Environment (DoE) created an Implementation Group of water company, environmental regulator and economic regulator representatives to coordinate UWWTD implementation in England and Wales and pass guidance onto their respective subaltern units.

However, another (bottom-up) stream in the implementation studies literature points out that these assumptions ignore ‘the reality of policy modification or distortion at the hands of policy implementers’ (Younis and Davidson, 1990, p.12). Top-level policy-makers in Member States may not wish to implement the Directive effectively. Even if they have the will, they may not be able to translate Directive requirements into detailed instructions for each organisational link in the implementation chain (Sabatier, 1986). Some discretion is inevitable. Moreover, the Directive requirements themselves may be somewhat vague and open to interpretation. Whether it is the top-level policy-makers or street-level bureaucrats, such discretion allows the Directive to be interpreted differently in the light of their contrasting prior interests and beliefs, or simply to cope with working pressures on the ground or to suit the prevailing political climate (Lipsky, 1982; Fineman, 1998). In effect, precise policy requirements only become apparent with hindsight, after they are translated into action, and are not necessarily clear a priori (Barrett and Fudge, 1981, p.25). Finally, some important resources may be possessed by actors beyond the direct authority and control of top-level policy-makers (Hanf, 1982), and such resource interdependencies afford opportunities for the resource possessor to bargain over the way a Directive is interpreted (Wilks and Wright, 1987, pp.4-5). In other words, the bottom-up view argues the operation of domestic policy networks beyond national administrations can significantly influence Directive implementation (see below). Implementation possesses important political dimensions (Barrett and Fudge, 1981; Ham and Hill, 1993; John, 1998). It implies this analysis of UWWTD will have to widen its scope beyond administration in order to explain responses to adaptive pressure.

A second process can subsequently be added to the fit administrative adaptation idea for further development:

1. the original adaptation pressure exerted by Directive requirement (however vague) upon a national administration (i.e. the top-down influence); and
2. the influence of interested national and sub-national policy actors upon interpretation of the Directive (i.e. the bottom-up influence).

Situations of moderate fit involve a confluence of Directive imperative and national influence, and anticipate a process of co-adaptation or mutual adjustment. The Directive presses a degree of change upon existing administrative practices, whilst domestic policy actors will seek to exert influence over the precise interpretation of the Directive and change in practice (Barrett and Fudge, 1981, p.24). The revised hypothesis is that the inevitable discretion in Directive requirement allows domestic interests and policy beliefs to influence national implementation of a Directive, subject to associated resource interdependencies. In this regard, whilst Knill does not use policy network terminology, his discussion nevertheless implies that fit can be influenced by domestic network processes between state and non-state actors, since these influence administrative change and adaptation (Knill, 1998, pp.4-9; O’Riordan and Jordan, 1996, p.76). Networks of actors clustered around key government departments - cemented together by resource interdependencies - include less formal arrangements that effectively mediate institutionalised activities and implementation processes (e.g. shared assumptions, tacit agreements to consult one another to the exclusion of others, common values, and rules of interaction) (Marsh, 1998).

From this broader perspective on implementation, the ability to adapt to a Directive depends upon the capabilities and resources of actors in the policy network (e.g. information, organisational capacity, economic resources, technical capabilities, authority to take decisions, legitimacy and support from key constituencies). So, for example, in Spain there was an absence of wastewater administration ahead of the UWWTD. European cohesion funds have provided the financial resources to build new treatment plant, but not the revenues to operate them. The success (or otherwise) with which the UWWTD has been implemented has depended upon the ability of the regional Comunidades Autonomas and local municipalities to build the organisational resources to begin collecting water charges to run the plant. This has posed a considerable political challenge in a country where water charges have been very low or almost non-existent (Calleja, Rebollo and Hemmelskamp, 1999).

Indeed, the willingness to adapt depends upon how adaptive pressure is perceived by network members in the context of existing rules of the game and the dominant policy beliefs in the network. Implementation will be effective if the Directive works with the national grain of core rules, procedures and policy beliefs, and only requires adaptation within them - rather than confronting the core and demanding wholescale changes. Depending upon the degree to which beliefs are shared within a policy network, and how these chime with Directive objectives, the ability to implement effectively may actually be in tension with the willingness to do so. Thus we might expect domestic networks toward the policy community end of the spectrum (i.e. highly cohesive networks of policy actors sharing common policy beliefs and exchanging influential resources) to be able to respond to a Directive (positively or negatively) much more readily than if the sector is inhabited by a much less cohesive issue network (i.e. few shared policy beliefs and poorly resourced) (see Rhodes and Marsh, 1992b; Bressers and O’Toole, 1994; Peterson, 1995). However, as has been argued elsewhere, it is understanding the operation of policy networks that is more important analytically than simply labelling them (Smith, 2000).

Fit, then, is no static lore. Its dynamism will mirror the extent to which the policy network changes over time. The Directive itself can demand change to networks, but so too can wider societal forces (Knill, 1998, p.8). How might these wider forces affect Directive outputs? Sometimes they may work to align the network closer with Directive demands, but other times they might exacerbate divergence. Coalitions of policy actors at odds with the domestic policy network, and by definition its core rules and beliefs, may, for example, demand change which supports or even surpasses a challenging Directive (Sabatier, 1998; Hajer, 1995). Changing socio-economic circumstances can also bring forth change (Smith, 1993). This suggests analysis of Directive implementation needs to be longitudinal and take heed of wider policy beliefs and socio-economic factors that impinge upon policy networks since these might alter the translation of a Directive into action. A conclusion similar to that from the implementation literature is thus reached, the national understanding of what the Directive requires will be shaped by domestic circumstance.

In summary, the approach proposed here stresses a need to move beyond a static comparison between Directive requirement and the immediate administrative tradition in place in a Member State (e.g. national laws and official guidelines). The theoretical discussion argues that fit, whilst usefully identifying adaptive pressure, fails to explain how adaptation might unfold. The UWWTD analysis needs to test this by identifying areas of discretion in the Directive and examining how they are used. The hypothesis developed above argues such areas will be interpreted to accommodate domestic policy network priorities, but that these can themselves be unstable and subject to wider domestic pressures. So we need to step outside the administrative unit responsible for delivering the Directive in the sector. Implementation can, arguably, be understood better through analysis of the interoganisational dynamics in the domestic policy sector (Knill, 1998; see also Hanf, 1982). Such an analysis needs to identify in these interactions the resource interdependencies, dominant beliefs, shared assumptions and rules of the game which underpin them (O’Riordan and Jordan, 1996, p.78). We need to analyse operations within policy networks and dynamic interaction with outsider actors and coalitions in order to understand better why some actors will be able to influence UWWTD interpretation, and consequently policy outputs, and not others. Here we might expect actors with resources important to UWWTD implementation to be most influential.

The structure of the paper tests this argument by broadly mirroring the approach suggested and assessing its explanatory adequacy; that is it takes a progressively wider survey of events surrounding UWWTD implementation. Empirical analysis begins by setting the national scene for implementation analysis. The English and Welsh water sector was privatised in 1989, and the evolution of a post-privatisation landscape was contemporary to UWWTD implementation. The section after that introduces the UWWTD in greater detail and applies a fit approach to implementation analysis in order to assess what can and cannot be explained this way. The adaptive pressure exerted by UWWTD objectives is assessed, and the administrative response is discussed (which was the creation of an Implementation Group by the DoE). In other words, these empirical sections test the fit approach to implementation analysis (see above). Subsequent analysis tests the alternative hypothesis developed above, i.e. that interpretation of Directive requirement and implementation outputs are shaped by the dynamics of the domestic water policy network under pressure from actors orbiting that network and responding to past events. The conclusion considers to what extent this wider approach enriches our understanding of UWWTD implementation in England and Wales.

The English and Welsh water sector

Privatisation, whatever the underlying complex of motivations (see Maloney and Richardson, 1995; Meredith, 1992, p.73), was marketed by government as the philosopher’s stone which could turn all the industry’s woes into successes. Private sector investment would be released and improve the infrastructure. This would be good for customers since private sector efficiency would ensure no concomitant increase in water charges (DoE, 1989). The initial reality failed to match the rhetoric. Charges were fixed by government in August 1989. They generously allowed for an average 5.5% real increase over the years 1989-94, and 5% for the period 1995-2000. This would fund £24.5 billion investment over the ten year period, including environmental quality obligations.

A study commissioned by the Department of the Environment (DoE), when the UWWTD was still a draft Directive in Brussels, anticipated UWWTD costs at £1.5 billion (House of Commons Environment Select Committee, HOCESC, 1990, p.xliii). The industry trade association, the Water Services Association, put it closer to £2.3 billion. Curiously, both estimates focused solely on coastal discharges. The assumption was that the UWWTD did not pose a challenge to inland STWs since these already treated to secondary levels and ‘would pass muster’ (Sharp, 1998, p.46). It was the coastal discharges that were under more immediate pressure from Europe, in the guise of legal action to implement the Bathing Waters Directive more robustly and criticism at North Sea conferences, and from domestic environmental NGOs (Jordan and Greenaway, 1998). Ministers were responding to these political pressures, and the £1.5 billion price tag appeared affordable in the context of the privatisation investment programme (Sharp, 1998, pp. 45-6).

The post-privatisation policy landscape evolved concurrently with UWWTD implementation. To prevent monopoly abuse by the ten regional water companies, a framework for economic and environmental regulation was created at privatisation. The Office of Water Services (OFWAT) was created for the independent economic regulation of the industry. OFWAT does this principally by regulating the charges each water company makes to customers. The National Rivers Authority (NRA) was created to administer environmental regulation, by issuing discharge consents, monitoring performance and enforcing compliance. It was replaced in 1996 by the Environment Agency (EA) - which incorporated the NRA and continued to operate along those lines. UWWTD implementation took place against this backdrop. The pattern of UWWTD implementation can be characterised as passing through two phases, which are analysed in the rest of this paper:

1. A minimalist implementation strategy, once the scale and cost of necessary improvements became apparent.

In 1993, during the first review of regulated water charges, the government sought to control escalating costs by limiting coastal treatment in as many cases as possible (through widespread ‘less sensitive area’ designation) and by minimising higher treatment inland with restricted designation of ‘sensitive areas’.

2. A revised approach which was more precautionary and uniform.

In 1998, during the second review of regulated water charges, the government tightened implementation of the UWWTD (by scrapping ‘less sensitive areas’ - consequently requiring improved treatment of coastal discharges) and increasing ‘sensitive area’ designations.

The changed approach suggests fit is pliable – Directive requirement can be reinterpreted, its demands are up for negotiation. What influenced the changed approach to UWWTD implementation? Subsequent sections explain this pattern of implementation. The account below is based upon corroborative analysis of primary and secondary documents and 29 elite interviews (e.g. people from water companies, economic and environmental regulators, government civil servants, environmental NGOs, trade associations, and wastewater equipment suppliers).

The Urban Waste Water Treatment Directive

The UWWTD is one of the more expensive pieces of European environmental legislation to reach UK shores and riverbanks. Reporting to the Commission in 1998, the UK government estimated total costs over the period 1993 to 2005 to be £8.9 billion (CEC, 1999, p.22). This is equivalent to £245 for every citizen of working age (16-65), although the revenue has ultimately come through household water charges. It contrasts with the original 1990 estimate of £1.5 billion. The struggle to contain escalating costs is an important component in explaining UWWTD implementation. High UWWTD costs have been influential in other Member States too. In Germany, high costs (especially in the new Länder where infrastructure is poor) have catalysed ongoing debates about the desirability of allowing the private sector to provide water services (Hansen, Kahlenborn and Kraemer, 1999). However, the English and Welsh case study presented here suggests narrow assumptions about capturing supposed private sector efficiencies ought to be treated very carefully.

Essentially, the UWWTD sets a timetable for meeting certain standards of discharge from sewage treatment works (STWs) - depending upon the population served by the STW, and the type of receiving water into which it discharges. The Directive distinguishes between three types of receiving water which, in terms of diminishing stringency of discharge limit, are Sensitive Areas (SA), Normal Areas (NA), and Less Sensitive Areas (LSA) (see Box 1 and Figure 1). The UWWTD requires these limits to be met using broad classes of treatment technology. The lowest limits can be met using primary treatment (physical settlement of contaminants), higher limits require secondary treatment (biological decontamination), and the highest limits need nutrient removal (tertiary) technology.

Box 1: statutory requirements and timetable under the UWWTD.

• All STWs greater than 10,000 population equivalent (p.e.) and discharging into Sensitive Areas to use advanced treatment to the following standards:

(a) Total phosphorous less than 2 mg/l (for 10,000 to 100,000 p.e.) and less than 1 mg/l (for p.e.>100,000), or an 80% reduction in total phosphorous;

(b) Total nitrogen less than 15 mg/l (for 10,000 to 100,000 p.e.) and less than 10 mg/l (for p.e.>100,000), or a 70-80% reduction in total nitrogen.

• All dumping of sewage sludge at sea to cease.

• All STWs greater than 15,000 p.e. and discharging to Normal Areas to use secondary treatment to the following standards:

(a) Biochemical oxygen demand (BOD) less than 25 mg/l or a 70-90% reduction in BOD;

(b) Chemical oxygen demand (COD) less than 125 mg/l or a 75% reduction in BOD;

(c) Total suspended solids (SS) less than 35 mg/l (for p.e.>10,000) and less than 60 mg/l (for p.e. between 2,000 and 10,000), or a 90% reduction in SS (for p.e.>10,000) and a 70% reduction in SS (for p.e. between 2,000 and 10,000).

• All STWs greater than 15,000 p.e. and discharging to Less Sensitive Areas to use primary treatment and need not meet standards as stringent as for discharges to Normal Areas, provided primary treatment only does not adversely effect the environment.

• All STWs less than 10,000 p.e. and discharging to Sensitive Areas to use secondary or appropriate treatment.
• All STWs less than 15,000 p.e. and discharging to Normal Areas to use secondary or appropriate treatment to meet BOD, COD and SS standards elaborated above.
• All STWs less than 15,000 p.e. and discharging to Less Sensitive Areas to use primary or appropriate treatment, provided treatment does not adversely effect the environment.

The Directive also has a general requirement for all sewage collection systems (including Combined Sewer Overflows) to be designed, operated and maintained using the best technical knowledge available not entailing excessive cost.

  Figure 1  

Adaptive pressure

The fit orientation suggests UWWTD implementation in England and Wales can be understood by assessing how the above demands match up to national administrative tradition. At face value, the UWWTD has relatively clear and precise objectives and leaves little flexibility over application. Its basic nature is a hierarchical, uniform emission standards approach within broad classes of discharge. That is, discharge standards are written into the Directive and serve as its point of departure. This is in contrast to the discretionary, environmental quality based approach traditionally practised in England and Wales (Hawkins, 1984; Maloney and Richardson, 1994; Haigh, 1989, p.22). This contrasting approach takes the receiving environment as its point of departure and is flexible over requisite discharge limits (NRA, 1994). The poor fit between UWWTD and national administrative tradition implies strong adaptive pressure would be generated during implementation. However, the Directive is vague on criteria for classifying receiving waters as SA, NA or LSA. It is left to Member States to set criteria for whether a receiving water is eutrophic (and therefore sensitive) or highly dispersive (and therefore less sensitive). Such discretion opens implementation to some of the bottom-up influences identified earlier. In England and Wales, this discretion has proved an important bargaining space during implementation. Costs have dominated debates. The two issues are connected in the sense that the type of receiving water sets the level of treatment needed, and this has obvious consequences for costs (see later).

Water infrastructure had suffered chronic underinvestment in the public sector, which plummeted from a high of around £950 million in 1973-74 (in 1986 prices, which is £1.5 billion in 1998 prices) to less than one third that level by the early 1980s, where it stayed until privatisation in 1989 (DoE, 1986). Unsurprisingly, underinvestment created performance problems for the wastewater treatment system. Since the 1950s, the water industry had, ostensibly, been designing and operating STWs to meet guidelines similar to standards required under the UWWTD (Cooper, 1998) (Box 1): all but the smallest inland STWs were practising some form of secondary treatment. In practice, however, many freshwater consents at privatisation did not include numeric standards, and a significant number of the works which did were failing their consent conditions (see Table 1) and/or had limits in excess of those in the UWWTD: 75% of works for suspended solids, and 61% for biochemical oxygen demand (BOD) (NRA, 1990). At coastal discharges there was little wastewater treatment prior to UWWTD implementation.

Source: Environment Agency (1998, p.47)

The UWWTD was consequently pressing for improvements in a number of areas, though the magnitude of change was not fully known at the time of signing (Sharp, 1998, p.45):

• poorly performing freshwater discharges (over 600 cases (HOCESC, 1998, p.xii));
• freshwater discharges to sensitive areas (a designation new to England and Wales);
• coastal and estuarine discharges (68% of the 404 discharges serving conurbations greater than 10,000 p.e. received no or only preliminary treatment (ENDS, 1990));
• improvements to 7000 unsatisfactory combined sewer overflows (CSOs); and
• a ban on dumping sludge at sea after 1998.

Poor fit extended to the state of English and Welsh infrastructure as well as approaches to regulating its performance and prioritising improvements. The UWWTD presented considerable demands and contrasted with the national administrative tradition. Sadly, we can surmise little more from this exploration of fit other than expecting ineffective implementation due to high adaptive pressure. This view suggests UWWTD outputs will be influenced by the discretionary, environmental quality-based tradition in England and Wales, such that improvements in infrastructure will be incomplete and not completely meet Directive ambitions. However, even this view is problematic because it assumes fit is somehow preordained. The fuzziness around LSA, NA and SA derogations is an instance in which the Directive contains a degree of elasticity - their practical meaning is open to negotiation in Member States. Fit alone seems to provide an incomplete explanation of the precise geometry of UWWTD outputs and the influences shaping them. It cannot explain completely the decisions over LSAs and SAs.

Administrative change: the UWWTD Implementation Group

The DoE convened a technocratic, quadripartite group to oversee UWWTD implementation that included the new regulators OFWAT and NRA/EA. The group also included the water industry trade association, and had the task of negotiating how terms and requirements in the UWWTD would be translated into operational criteria and procedures. Technical, multi-party groups such as these are a common DoE mechanism for implementing Directives. Guidelines were issued to water companies and local NRA offices so that environmental managers and inspectors could implement UWWTD tasks consistently, through the review of existing discharge consents. The guidance had no statutory force, so a degree of consensus amongst all parties was essential if it was to become operational.

In some respects, the Implementation Group can be understood as an attempt by the DoE to forge an implementation chain. Whilst negotiating consensus interpretations of Directive terms was at times highly charged, the objective was nevertheless to manage implementation tasks by ensuring all administrative changes were agreed and relevant actors coordinated their efforts. On first inspection, the Implementation Group appears to be the location where adaptation to the Directive was negotiated. However, the Group did not take the decisions over LSA and SA designations. It was during the episodes of economic regulatory review – beyond the realms of the Implementation Group - that these important parameters for interpreting the UWWTD were really set. Economic regulatory reviews fixed the parameters for water policy generally, of which UWWTD implementation was a significant part.

How were decisions reached on whether to utilise LSA derogations and classify SAs? The precise requirements of the UWWTD (i.e. how it fitted into the national scene) were not immediately obvious, and needed clarification by DoE for the purposes of economic regulation (see below). These were important decisions that had widespread effects on UWWTD outputs across the sector. The quadripartite Implementation Group agreed technical routines for meeting various UWWTD requirements, but it was government guidance which decided whether to activate those requirements or not. Negotiations in the Implementation Group cannot explain the shifts in this guidance. Can resource interdependencies and policy beliefs in a wider post-privatisation policy network explain government guidance (and UWWTD outputs), as our second hyptohesis suggested? Theoretical discussion also pointed to policy beliefs and events beyond the policy network that influence its operation and hence the pattern of UWWTD implementation. Having illustrated the limits of a narrow fit approach, it is time to explore the explanatory effectiveness of this wider perspective. Subsequent sections broaden the analysis, and test the hypothesis that it is domestic policy beliefs and resource interdependencies that will explain UWWTD implementation.

Implementing the UWWTD - a broader view

The parameters within which the UWWTD has been implemented (and Implementation Group guidelines thereby activated) were set by events surrounding the Periodic Reviews of prices. OFWAT has conducted two Periodic Reviews of regulated water charges, first in 1994 (AMP 2) and again in 1999 (AMP 3). It is important to note that negotiating the Periodic Review operated at a more senior level than the technical experts of the Implementation Group, owing to its significance for water policy more generally. Nevertheless, deadlines for UWWTD implementation have overlapped with OFWAT periodic price reviews such that UWWTD improvements have become an important component in investment plans (see figure 2).

Figure 2: overlap between the UWWTD and economic regulatory timetables.

Ultimately, it was the water companies who held the organisational and financial resources to make necessary UWWTD improvements. Each Periodic Review anticipates the investment they require over the coming ten years. At the same time, OFWAT is keen to ensure water prices are kept as low as possible for customers, and so has an interest in ensuring the necessary investments will be made efficiently. It is supported in this cost-averse policy belief by consumer representative groups, such as the ten regional Customer Service Committees, which advise OFWAT through its National Customer Council (ONCC, 1998; Reiter, 1998). This positions OFWAT as an important player in the shaping of water quality strategies (Rees and Williams, 1993).

The first Periodic Review: setting the parameters for UWWTD implementation

The first Periodic Review in 1994 sparked a panic over the cost of the UWWTD, which seemed to be spiralling well beyond the initial £1.5-2.3 billion estimates. The revised estimates, though contested, were nevertheless on a rising trend. A cost containment strategy sought a minimalist interpretation of UWWTD demands. The NRA/EA then went about its implementation task (i.e. revising discharge consents) within these minimalist parameters, following the relevant guidelines drawn up by the technical Implementation Group.

OFWAT’s 1992 projection for the UWWTD was a qualified £6 billion: ‘At an early stage of planning margins of error are wide. Previous experience suggests that costs are more likely to increase than decrease. Margins of error will tend to be skewed; final costs could easily be half as much again or above those in this paper’ (OFWAT, 1992, p.13). In July 1993, the Director General of OFWAT, Ian Byatt, told Ministers he needed ‘clear guidance from them on the new [environmental] obligations that companies are likely to face and the time scales for achieving them. When these obligations are established I have a legal duty to ensure that they can be financed’ (OFWAT, 1993, p.4). Here we have an explicit example of a national policy actor for whom officially sanctioned UWWTD requirements are far from obvious. In making this call, Byatt presented revised estimates for his interpretation of UWWTD demands, whose ‘broad totals are sufficiently robust for decisions to be taken on the scale of new obligations’ (1993, p.4). OFWAT now thought the UWWTD would cost £10 billion.

One of Byatt’s stated objectives for the Review was to protect customers from excessive price rises (OFWAT, 1992, p.7), which had been increasing in real terms by 5% each year. Every £1 billion capital expenditure added around £3 a year to average household water charges; and every £100 million in operating costs had the same effect (OFWAT, 1993, p.16). Whilst Byatt sought UWWTD guidance from Ministers, he remained very mindful of impacts on customer water charges and was at pains to advance the cost-averse policy belief:

I believe that the Secretaries of State should consider going as far as renegotiation of European standards, where obligations, and particularly the speed of their implementation, place an unmanageable burden on customers. In interpreting EC Directives I believe that account should always be taken of the impact on customers’ bills, in order to achieve the most timely and cost effective solutions’ (OFWAT, 1993, p.13).

Rising UWWTD estimates were a device for alerting high-level political interest in household water charges and recruiting political and authority resources to the cause of cost-averse policy beliefs.

Meanwhile, an ad hoc programme of freshwater STW consent reviews had begun after privatisation, as part of plans for meeting river quality objectives drawn up by the NRA, which was keen to promote an image as ‘guardian of the water environment’ (NRA, 1995). Area offices of the successor EA have been introducing UWWTD requirements into this broader review of STW consents. However, river quality objectives are not statutory requirements, nor is there an improvement timetable, and this limits the force with which the NRA/EA can demand STW improvements. Only statutory limits, such as those under the UWWTD, or objectives with explicit government backing can be really pressed, since it is these that are recognised under the economic framework and are incorporated into investment plans and price controls.

The NRA thus had an incentive to maximise requirements interpreted under the UWWTD since these are, an admittedly blunt, means to higher river quality objectives. OFWAT, conscious of the effect on water prices of any environmental investment, sought to limit the NRA’s demands upon water companies only to those objectives required by statute, which should be based upon strict cost-benefit considerations (OFWAT, 1993, p.13). The two regulators thus have different interests and hold contrasting policy beliefs.

So drawing the boundary between which environmental improvements could and could not be legitimately demanded under the UWWTD concerned both the NRA and OFWAT. In effect, the authority resources of the NRA were limited by what were deemed legitimate and fundable improvements under the economic framework. Yet the identification of necessary improvements, needed by OFWAT for cost estimates, would be done by the NRA. There is a structure of resource interdependency here that requires agreement over objectives, including common interpretation of UWWTD requirements. Striking the balance between the maximalist environment regulator and the minimalist economic regulator was the DoE (since 1997, the Department of the Environment, Transport and the Regions (DETR)) (see figure 3). The DoE was still learning how best to perform this role as regulator of regulators during UWWTD implementation (DoE, 1991, p.10). A nascent post-privatisation policy network closer to the issue network end of the policy network spectrum emerges from this analysis.

Figure 3

The government had to strike the balance between competing interests in the water sector. It was also under political pressure from a variety of actors beyond the policy network over its performance (i.e. the way OFWAT was regulating prices and the effectiveness with which the NRA was delivering improvements) (figure 4).

figure 4

Greater transparency under the post-privatisation policy landscape had made it easier for outsider groups to monitor the sector and have a voice (Maloney and Richardson, 1995, p.160). This made the post-privatisation policy network that much more susceptible to outsider intervention (another characteristic of issue networks). Environmental NGOs, Parliamentary Select Committees, local communities, and consumer groups were all transmitting their policy beliefs to government, which had to decide how best to guide UWWTD implementation (as well as water policy more generally). As we shall see, these policy beliefs essentially amplified and extended those held by OFWAT and the EA. That is, they fell into two broad camps: a cost-averse policy belief; and an environment-oriented policy belief.

Growing unease over two aspects of water company performance were vying for government attention. There were continued campaigns against their environmental performance by groups like Friends of the Earth and the newly created, and highly media effective, Surfers Against Sewage. Nothing short of full tertiary treatment would be sufficient for these groups (SAS, 1992, pp.17-18). In addition, profits and large executive pay rises against a backdrop of rising water charges were fuelling public disquiet about the economic performance of water companies. The South West was a particular crucible for allegations about profiteering in the context of very steep increases in water charges. Opposition M.P.s saw political capital in fanning the flames of discontent, to the alarm of sitting Conservative M.P.s (House of Commons Debates, 31st March 1994, Vol.240, Col.1093; ENDS, 1994, p.6; Summerton, 1998, p.62).

OFWAT’s alarming conclusions over UWWTD costs coupled with Parliamentary disquiet were sufficient to draw Treasury and even Prime Ministerial intervention - neither renowned for a keen interest in water policy. Both applied pressure for UWWTD implementation to be managed as ‘cost-effectively’ as possible (House of Commons Debates, 21st June 1994, Vol.245, Col.122). Cost control became paramount and followed two paths: one led to Europe, the other pressed high-level political concerns upon domestic UWWTD interpretation. The European strategy sent senior DoE civil servants to their counterparts in other Member States ‘to explore the possibility of revising the deadlines and other requirements of the UWWT directive: this has required time-consuming discussions at a senior level in Brussels and other European capitals’ (DoE, 1994, p.7). The Chancellor, Ken Clarke, brought the Treasury into the fray and exhorted other EC Finance Ministers to back the UK in revising the UWWTD (ENDS, 1993, p.16). Unfortunately for the government, these discussions with European counterparts did not bear fruit, even though Germany has subsequently made similar noises.

The second cost mitigation path was to take full advantage of LSA derogations, which the government termed High Natural Dispersion Areas, whilst a more cautious approach was taken over the designation of SAs (DoE, 1993, p.7). This minimalist implementation strategy was conveyed to regulators in DoE guidance (DoE, 1993, p.9). Long stretches of the UK coastline were designated under 58 LSAs in May 1994 (DoE, 1994b). Improvements to 6000 failing CSOs were staggered over twenty years. As for SAs, 20 SAs were designated and decisions on a further 29 (which had been recommended by the NRA) were deferred by government and subject to further monitoring (DoE, 1994b). The number of coastal discharges needing to invest in secondary treatment would thus be held in check, and nutrient removal from freshwater discharges likewise (Haddon, 1998, p.16). This designation contrasts with Nature Conservancy Council and English Nature studies, which found a far higher incidence of eutrophic waters effected by sewage (ENDS, 1992, p.36; Haigh, 1998, 4.6-9).

The minimalist approach was not unique to UWWTD guidance and, in retrospect, not that surprising. The government had been similarly lacklustre in its approach to designating beaches under the Bathing Waters Directive and fisheries under the Shellfish Waters Directive earlier in the 1980s. The government was more likely to be sympathetic to the cost-averse camp than the environment-oriented demands, particularly when it became apparent that the UWWTD would be more expensive than expected. As soon as costs started to rise, all the signs of the positive response to environmental pressure that had led to the government signing the UWWTD in 1991 simply disappeared. OFWAT and not the NRA got the guidance it wanted with the favourable deployment of government authority resources - even though the legitimating effect of the environmental-quality based tradition (i.e. dilute and disperse) was by now wearing thin.

The final outcome of the Periodic Review anticipated costs for the Directive to be £6 billion over the following ten years (OFWAT, 1994). This is a similar level of investment to that in France, where implementation has been far less controversial, perhaps because the vast majority of funds came from general taxation rather than sharply escalating water charges (Jordan, Ward and Buller, 1998, pp.1490 & 1403). With planned costs in check in England and Wales, real increases in water charges (the K factor) were limited on average to 1.5% each year. UWWTD implementation had taken its first twist.

The first Periodic Review taught the economic and environmental regulators some important lessons about the pliable nature of UWWTD demands. The two regulators also learnt that the government was an important holder of the ring. The NRA concluded it would have to make ‘strong representation to government’ in the future if it was to meet environmental quality ambitions (NRA, 1994b, p.14; ENDS, 1994, p.6). OFWAT was determined to seek cuts in water charges in the second Periodic Review five years hence.

Explanation of this phase of UWWTD implementation certainly appears to benefit from a wider analytical perspective. The machinations of the post-privatisation policy network, particularly economic and environmental regulatory interaction, forced cost considerations upon the government. A lack of cohesion and the nature of interdependencies in the policy network facilitated the intervention of government, who did so with terms attached. Actors and events beyond the policy network sought to influence government guidance over UWWTD interpretations. Some actors were directly interested (e.g. English Nature) whilst others had more diffuse concerns which fuelled the decision-making climate indirectly (e.g. opposition M.P.s). The historically stronger cost-averse policy belief remained more influential than the environment-oriented policy belief. The use of LSA derogations and restrictive SA criteria reflects this lack of environmental willingness at the top. The legitimating persistence of UK quality-based regulatory traditions might have made this approach palatable to the NRA had the financial savings been redirected to other river quality objectives, but they were not. Second time around the economic regulatory cycle the environmental regulator would press for tighter UWWTD interpretation.

The significant point arising from this episode is how, with both the unsuccessful attempt by government to go back to Europe and seek modifications to the UWWTD, and with a minimalist interpretation of the existing requirements, government actively sought to manage Directive requirement in a highly political implementation context.

The second Periodic Review: a more precautionary approach

Although details of many of the subsequent implementation tasks were done by the NRA and water companies away from the public gaze, the outputs were not without controversy. There was concern over two LSA decisions in particular (in the Bristol Channel and the Humber Estuary) and continued pressure from environmentalists for higher levels of treatment for coastal discharges in general. The political saliency of these issues generated conflicts that were to shape the second Periodic Review, launched in 1997. Stoking the politics over environmental quality were continuing profit levels in the industry, perceived to be excessive by critics (including the opposition Labour Party). Interdependencies between the economic and environmental regulatory frameworks brought the environment-cost debate sharply into focus, which influenced the second shift in UWWTD goalposts.

The Environmental Debate

Controversy over LSAs detonated when Environment Secretary John Gummer set the boundaries for estuarine and coastal waters in the Humber and the Bristol Channel. He chose them such that two large discharges (from Hull and Bristol) fell on the favourable, minimalist side of the boundary, where coastal LSA status suggested they would not require secondary treatment. The boundaries chosen in these cases - road-bridges across the stretch of water - were inconsistent with the hydrological boundary criteria used for all other estuaries. Local authorities challenged the decision, and the government found itself in the High Court. The Court judged that the government had failed to apply ‘a genuine and rational assessment’ and it was bending UWWTD interpretation too far (ENDS, 1996). This decision did not cast the minimalist approach in a favourable light - its legitimacy was undermined. It also added £120 million worth of secondary treatment (and £5 million annual operating costs) to the Hull and Bristol upgrades.

Even if all Comprehensive Studies showed that primary treatment for discharges into LSAs was sufficient, it did not absolve all coastal discharges from more costly secondary treatment. Bathing water quality standards were requiring some companies, but by no means all, to install secondary treatment at relevant STWs. In a few cases, water companies were also experiencing trenchant local and environmentalist criticism of plans which limited upgrades to primary treatment. With the encouragement of high media profile campaigns from SAS, groups in some areas were even demanding the installation of tertiary treatment, such as UV disinfection. One such location was Aberystwyth, where Welsh Water faced sustained opposition. Nor was the company particularly confident that LSA status would hold forever. Local difficulties and doubts prompted a more precautionary interpretation of the UWWTD in this particular case, and to look at the costs and feasibility of UV disinfection. They concluded that full-treatment was an affordable option and they decided to install UV (Gibson, 1992).

By breaking rank the company had set a practical precedent that won valuable public support from SAS and other non-network members, and boosted their environment-oriented arguments. If Welsh Water could afford to invest beyond statutory compliance (i.e. primary treatment), then why couldn’t the other water companies? In time, a few others did commit to full-treatment at some STWs (Haddon, 1997, p.17; SAS, 1998, p.11). Environment-oriented policy beliefs were boosted further with the recruitment of the House of Commons Environment Select Committee (HOCESC) in February 1998, just as the debate over OFWAT’s second Periodic Review of prices was getting into full swing (see below). HOCESC recommended all sewage should be treated to tertiary level at all times and in all places (HOCESC, 1998, pp.lxxii-lxxiii).

This pressure for higher treatment was beginning to bite, and the newly elected Labour government (May 1997) responded to HOCESC with a partial revision to the UWWTD approach. Comprehensive Studies for 12 coastal discharges had failed to demonstrate that primary treatment was sufficient and, with encouragement from the EA, the government decided to revoke LSA status at 21 of the 58 LSAs. The government hinted that more might follow, stating that it ‘also wishes to consider very carefully whether, notwithstanding the scientific evidence offered by the comprehensive studies, a precautionary approach should dictate that secondary treatment should always be required for significant coastal discharges’ (DETR, 1998, p.5, emphasis added).

The government decision did not meet all environment-oriented demands, but it nevertheless significantly modified the fit between UWWTD and national administration. Moreover, in the same response the government announced the designation of a further 47 SAs. A degree of uncertainty over future UWWTD demands had been introduced. However, advances in environment-oriented argument beyond the policy network provides only a partial explanation for the government’s stricter reinterpretation of the UWWTD. The new government remained conscious of cost and political implications; so why had they decided upon tougher UWWTD steps where their predecessor had feared to tread? Unlike their predecessor, the new government had the benefit of hindsight over outputs from the first Periodic Review, which had done little to dampen a continued debate about profiteering in the industry.

The Profitability Debate

Water charges continued to rise after 1994, as allowed under the first Periodic Review. In 1986-87 average household water and sewage charges had been around £145 (in 1998 prices) (HOCESC, 1987, p.459; OFWAT, 1992, p.9). By 1998-99 they were just under £250 (OFWAT, 1998, p.24). Some of the revenue from this 72% real increase in prices was funding UWWTD implementation. However, the revenue was also generating profits perceived by critics as excessive. Within a year of the first Periodic Review water companies had announced share buy-backs and dividends totalling £1.5 billion (ENDS, 1997), and over £1 billion of dividends were paid out each year from 1994 to 1998 (OFWAT, 1997b, p.26). A greedy public image problem was exacerbated by fat cat pay rises to senior managers (Daily Telegraph, 21/12/94; Summerton, 1998, p.61).

In opposition, the Labour Party had milked public dissatisfaction over water profits for all its political worth. Utility companies were a soft and popular political target. In government, Labour introduced a one-off windfall tax on the privatised utilities, including £1.1 billion from the water companies, which would fund their youth training programme – the New Deal for the Unemployed (Labour Party, 1997, p.19). Moreover, environment portfolios were the domain of John Prescott and Michael Meacher - both from the traditional left and unlikely to take so sympathetic a line with water companies.

The scale of profit making suggested OFWAT had miscalculated the first Periodic Review (Summerton, 1998, p.63). Regulation of prices is intended to provide incentives for (financial) efficiency and innovation, since any cost savings beneath fixed prices count towards enhanced profits for the companies (OFWAT, 1997). In practice, the water companies also have an incentive to inflate their planned capital expenditure and exaggerate their operating costs such that OFWAT sets a generously high price limit in the first place - a profit boosting tendency termed ‘gold plating’ in the industry (Helm, 1994, p.21). OFWAT, dependent upon industry information, had anticipated operating expenditure rising by 3% over the AMP2 period - in the event it fell by 12%. Capital cost savings on STW upgrades were 30-40% greater than expected (ENDS, 1999, p.42; OFWAT, 1998b, p.16). If all UWWTD related work could achieve this efficiency then the £6 billion capital expenditure allowed in AMP2 would actually cost the companies £3.6-4.2 billion. Clearly, the UWWTD was not costing the amounts allowed under AMP2. Hindsight indicated there was financial room for tighter UWWTD implementation, or cuts in water charges to customers, or both. It was this issue of distributing benefits that became the focus for the second Periodic Review, launched in June 1997.

Policy actors were pressing points or disseminating interpretations of information which, they hoped, would reinforce the dominant cost-averse policy belief or see it replaced by an environment-oriented belief. This applied to water policy in general, not simply to the UWWTD, yet clearly whichever view prevailed amongst Ministers would set guidance on the latter issued under the second Periodic Review.

Moving the UWWTD Implementation Goalposts

At times the Review was acrimonious, with both OFWAT and the EA making public declarations about their priorities, pointing to what they believed to be weaknesses in the other’s analysis, and actively lobbying the government (EA, 1998; OFWAT, 1998c; ENDS, 1998). As with the first Periodic Review, it was the government that would effectively strike the balance between environmental quality and customer water charges, since it was they who would guide OFWAT and the EA on the UWWTD investments deemed to be necessary over the coming AMP3 period.

Early in 1998 the industry trade association, DETR, the EA and the National Customer Council all conducted polls to measure public opinion about water charges and the environment. It is not the precise outcomes of these polls that is important here: they were all framed differently, used varying sample sizes (from 48 to 2489), and asked different questions. What is significant is the fact that the opinion polling exercises took place. It reveals the political nature of the Periodic Review and that policy actors realised the significance of impressing upon the government a certain view of what was in ‘the public interest’. The economic and environmental regulators spent the summer of 1998 jostling for positions of influence over the forthcoming guidance, as did network non-members.

The government finally issued guidance to OFWAT in September 1998. It went even further than their response to the HOCESC inquiry the previous year. They decided to take the more precautionary approach by revoking all LSAs and requiring secondary treatment for all significant coastal discharges (DETR, 1998b, p.19). The CSO programme was accelerated (DETR, 1998, p.21). These requests formed part of a package of environmental investments which, given evidence from the profitability debate, the government believed could be met without price rises (DETR, 1998b, p.3).

The Periodic Review continued and price cuts were announced by OFWAT in November 1999. However, the impact of the Review, so far as implementation of the UWWTD was concerned, had already been felt. The resource interdependencies remained the same in this second Periodic Review as they had been in the first. However, this time around the wider politics were different. Not only was there a new government, but events in the interim had belied some of the water industry and OFWAT cost projections and strengthened the position of their opponents. Emboldened by past efficiencies, the government felt it feasible to tighten interpretation of UWWTD requirements. It was in the fortunate position of being able to answer some of the environment-oriented demands without upsetting cost-averse policy beliefs. This explains why the fit between the UWWTD and domestic practice passed through a second configuration, with obvious consequences for outputs. Thus the latest reinterpretation of UWWTD requirements has further advanced the performance of the English and Welsh wastewater infrastructure. The experience of this ratcheting implementation process suggests a few more twists and turns in the future. It also shows that England and Wales has not performed particularly well over the UWWTD, in the sense that households have paid more than was necessary for treatment that still falls short of the best available. This environmental performance lags behind the Netherlands and Germany, where tertiary treatment is being installed as the norm, and reflects the mid-league position of Britain in relation to European environmental policies generally (Lowe and Ward, 1998).

Discussion and conclusions

The fit between the requirements in a Directive and Member State administrative tradition provides policy analysts with an approximate measure of adaptive pressure. The fit approach hypothesises that the greater this pressure the less effective implementation is likely to be. This paper has sought to argue, theoretically and through empirical application, that this fails to explain the ‘post-decisional politics’ that shapes implementation outputs and outcomes (Puchala, 1975, cited in Jordan, Ward and Buller, 1998, p.1389). Precise requirements in the UWWTD have been shown to be far from obvious at the outset - contrary to what is implied in the fit perspective. Indeed UWWTD requirements have been subject to intense political debate in England and Wales. Fit is pliable, not preordained. In this respect, UWWTD implementation in England and Wales has been a particular instance of a more general fight over domestic water policy priorities. Moreover, whilst it is true that the modus operandi for the UWWTD contrasted with traditional practice in England and Wales, recognition of this fact tells us little about the way each has mutually adjusted to the other.

It has been a contention of this paper that such mutual adjustment, and the policy outputs this process generates, can be explained through the operation of domestic water policy networks and the influence of wider events upon them. An alternative hypothesis to that of fit was proposed. Namely, that resource interdependencies would draw different policy actors into the implementation process, and that the dominant policy belief would colour the interpretation of those Directive requirements over which there was discretion. In England and Wales resource interdependencies brought a number of actors into the fray over UWWTD, but there was little cohesion between them. Nevertheless, it is the resource interdependencies between OFWAT, the EA , the water companies and government that explain the involvement of each. A policy network approach to analysis does seem at least partially beneficial.

Implementation of the UWWTD was debated by an issue network that needed government intervention in order to fix Directive requirement. In turn, the government has been mindful of the political pressures upon it, be they cost-averse or environment-oriented policy beliefs. Explaining government guidance has required an analysis of the way these competing policy beliefs have garnered influence or lost it. Events beyond the immediate interactions between economic and environmental regulators add to the explanation of UWWTD implementation. There is also evidence of policy learning with the outcomes of past Directive interpretation informing revised government guidance (see also Barrett and Fudge, 1981, pp.24-29). The environment-oriented policy belief became influential once it became apparent to government that the former’s objectives could be met (at least partially) without upsetting more entrenched cost-averse policy beliefs.

It seems fair then to conclude that in cases where an issue network is responsible for Directive implementation its requirements will be susceptible to reinterpretation, owing to its penetrability, and that outputs will shift under events and activities beyond the immediate implementation tasks. This is different to simply saying implementation will be ineffective, since throughout the English and Welsh UWWTD experience implementation has conformed with the Directive (apart from Bristol and Hull exceptions). It is the official understanding of what conformance requires, and hence performance, that has changed. In England and Wales that understanding has been highly contested compared to the situation in countries with cohesive policy networks such as the Netherlands, or France (Jordan, Ward and Buller, 1998, p.1404). The way Member States understand and perform against a Directive will reflect the operation of their policy networks and how these mediate domestic political concerns.

Acknowledgements

Many thanks to Andrew Jordan and two anonymous referees for helpful comments on an earlier draft of this paper. Thanks also to the editors for their prompt assistance.

References