Decentralized reflexive environmental regulation: Opportunities and risks based on an evaluation of Dutch experiments

Abstract

The main tenet of reflexive law theory is that, to solve environmental problems, procedural regulations may be more effective than detailed and substantive prescriptions for action. This paper explores the viability of this tenet by analysing two experiments in urban environmental and spatial planning whereby classic substantive environmental regulation is replaced by a form of reflexive regulation, enabling policy-makers to provide a flexible approach to achieving environmental improvements. The analysis confirms, though only conditionally, the viability of reflexive law theory. Important conditions are the definition of environmental quality, the organization of participatory processes, an external agency to validate the decision-making process, and cooperation of planning agencies. It is also observed that reflexive law tends to neglect the underlying societal processes of environmental problems. Some societal processes are irreversible. They cannot be changed by classic regulation nor by reflexive law.

Keywords:

• Decentralized reflexive environmental regulation
• opportunities
• risks
• Dutch experiments

1. Introduction

Confronted with the complexity of environmental problems, and sustained by the current liberal political climate, many western governments are pro-actively seeking a new role to conduct their duty to protect the environment (Driessen and Glasbergen 2002).

A general feature of these endeavours is the move away from detailed and centralized norm-setting (government) towards decentralized procedural requirements for the processes within which a desirable environmental quality needs to be defined (governance). Problem-solving capacities are seen as often blocked by rules that are over-specified, objectives that are too detailed, and too little policy freedom for relevant actors, inhibiting their search for creative solutions. New forms of ‘reflexive’ regulation should be developed that facilitate the process of defining the environmental quality instead of prescribing this quality beforehand. This shift from government to governance brings in more actors into policy-making processes in an attempt to mobilize their problem-solving capacities. It also creates opportunities to define environmental objectives and norms adapted to the specific circumstances within which they need to be applied.

In the 1990s, the Dutch government introduced a form of reflexive regulation—a decentralized form of planning which enables planners to provide a flexible approach to achieving environmental improvements—into its environmental policy (VROM 2001a). In view of its objective to protect the environment, the central government took the stance that certain risks are unacceptable—risks entailing irreversible damage to people and the environment—and must be subject to strict regulation. Under certain conditions, however, some adverse environmental effects may be allowed. In this large grey area, the legal system need not specify what is or is not allowed; rather, it could create room for a re-evaluation of classic substantive environmental legislation and working it into more reflexive ones. These ventures were prompted by a sense that environmental law had become too specific and too strict. It was felt that strict, nationwide environmental norms could obstruct reasonable decision-making on the interests in environmental and spatial planning. A flexible response to complex environmental and spatial planning situations requires more scope for decision-making.

In the mean time, two experimental forms have been implemented. These experiments reframed the relationships between government actors, delegating more responsibility for solving environmental problems to the local authorities. One concerns the drafting of voluntary noise-abatement plans, giving local government greater freedom to define its objectives for a quieter environment. The other concerns an experimental law for collaborative inner-city redevelopment, a law that has given local authorities more say in setting environmental standards. Both offered reflexive alternatives to current substantive regulation and replaced strict nationwide norms with a broad concept of environmental quality. The experiments in reflexive environmental regulation entailed several innovations:

1.	a national normative framework, relaxing strict standards and creating opportunities to apply a broader definition of environmental quality;
2.	a set of procedures, regulating any deviation from national standards and streamlining communication between stakeholders; and
3.	a system to validate the decision-making process, giving the outcomes their political legitimacy.

These experiments relate to the German-based theory of reflexive law (Teubner 1983, 1997, Teubner et al. 1994, Willke 1997, Calliess 2001). This theory is both a descriptive and a normative one. In its descriptive sense, it aims to explain the mechanisms through which the legal system relates to other sub-systems in society. In its normative sense, it can be seen as promoting a multi-level approach to governance which depends, for its effectiveness, on decentralized forms of deliberation (Barnard et al. 2004: 2). The cases refer to this normative interpretation. According to this interpretation, societies might be better off with procedural or constitutive rules rather than strict substantive regulative rules. Constitutive rules are rules that induce stakeholders in an issue to reflect on their own actions; to adopt a critical view of their own operations and management. The presumed advantages of reflexive law have been summarized by Orts: it is less costly and more flexible; it can mobilize the technical skills and resources that a central government can hardly provide; and it can be tailored to suit those directly involved. This adaptive capacity allows for direct links between making the decisions and living with the consequences, thereby fostering a greater commitment towards implementation (Orts 1995). Some of these arguments underlie the Dutch experiments. The assumption of the experiments is that the local authorities are the best judges of the location-specific determinants of environmental and spatial problems, as well as the best judges of a suitable environmental quality—better than any other tier of government. Furthermore, the local authorities, being in direct contact with local stakeholders, are assumed to be in the best position to organize a participatory decision-making process.

While the theory poses that societies might be better off with contitutive rules rather than strict substantive regulations, it remains abstract, skirting the intricacies of practice. Little is still known about the design of effective reflexive policy-making processes. The aim of this paper is to explore the viability of a specific form of reflexive regulation, here called decentralized reflexive regulation, stimulating local government to reflect on its own orientations and policy practices and exemplified in the two Dutch experiments in urban environmental and spatial planning. Though different in scope, both models redefined the classic substantive regulatory approach and introduced innovative procedural regulations which constitute environmental norm-setting as a communicative process. The main questions are: what mechanisms and methodologies are introduced to induce reflexivity, and what are the effects in terms of improved problem-solving capacity? The answers to those questions reveal some critical conditions that need to be fulfilled to constitute a reflexive policy practice.

2 Case study 1: Reflexivity in regulating noise pollution

One of the first packages of environmental rules to come up for revision was the Dutch policy on noise pollution. The current law is complex and detailed; in the course of time, it has been amended to include norms for each source of noise (traffic, railways, industry, etc.). Each norm has an upper limit indicating the maximum allowable noise level and a lower limit indicating the preferable maximum. A specific method to calculate the noise emissions has been allocated to each source. Additional regulations stipulate how the norms should be applied spatially and which instruments should be used. Some rules specify options for dispensation, each one indicating the acceptable margins per source.

In view of such complexity, it is no surprise that the local authorities, being responsible for implementation of the rules, were among the first actors to look for alternative approaches to regulation. An analysis of the current law revealed high implementation costs, for both local government and private developers. It also indicated that insufficient opportunities had been created to find flexible solutions in practice. The analysis also showed that the exemption rules led to unnecessary delays in the spatial planning process, losses that were not offset by significant environmental gains. The investigators concluded that environmental norms hamper an integrated approach to planning for urban development. The reason is that the norms are not usually considered until the design process is in its final stage; by that point, they have overrun all margins (MDW-werkgroep Wet geluidhinder 1996). Another common effect of detailed regulation that showed up in the analysis was a continual need to adapt the rules in the face of new developments.

A new law was needed to create opportunities for practical solutions; it would have to shift responsibilities from the central government to the local level; and it would have to lower the transaction costs while preventing noise pollution (VROM 1998). Taking these ideas as the point of departure, a new piece of legislation was prepared (VROM 2001b). A key feature of the bill was the shift from a norm-setting system towards a system of area-related objectives. The bill set an absolute nationwide upper limit of 70 dB that could not be exceeded under any circumstances for reasons of public health and a preferable maximum of 50 dB. Both limits reflect cumulative acoustic effects outdoors; this means that the total amount of noise from all sources is decisive. A more specific standard was set for noise levels measured indoors.

The local authorities were supposed to decide on any deviation from the maximum preferable noise level and lay it down in a voluntary noise abatement plan. In the absence of such a plan, the preferable level serves as the absolute maximum, and deviations are no longer permitted. The bill advises the local authorities to inform their inhabitants about the plan in an interactive process, though the bill does not explain how to do this. Local governments have to include the emissions of noise by regional and national infrastructure (i.e. railways and highways lying outside their territory) in their plans.

The point of departure for the bill is one of the basic assumptions of reflexive law theory. It is based on the premise that regulations are most effective when they originate in the system that is to be changed. Indeed, most instances of noise nuisance are caused by local circumstances. One might expect local governments to have better insight into the location-specific determinants and the consequences than any other tier of government. Since they are also the main decision-makers in spatial planning and urban design, local administrators are able to redirect sources of noise pollution to other sites. These new rules were also expected to transform noise abatement from a purely technical problem into a social issue, thereby giving it a more prominent place on the local political agenda.

2.1 Results of the pilot projects

The central steering mechanisms in this case are voluntary noise-abatement plans. To gain experience with the bill, seven local governments started to develop such a plan as a pilot project in 1999. The central government made some recommendations beforehand. For instance, the plan should set forth area-specific local goals for preventing and reducing noise nuisance, and indicate how these goals should be achieved. Some methodological innovations were introduced as well. One new feature was the opportunity to treat noise as a cumulative problem, combining the contribution of all sources in one aggregate standard. Another was that the goals could be translated into area-specific quality targets. Doing so would allow the authorities to abandon the approach solely focused on noise abatement and to set targets with reference to the quality of the living environment. A third innovation was the strong recommendation to involve local stakeholders in the process of policy-making.

To assess the effects of this kind of reflexive regulation, all of the pilot projects were analysed. Next, to document analysis, all major public and private stakeholders in the seven selected cases were interviewed. The analysis was focused on three topics:

1.	goals of the noise-abatement policies;
2.	participation in the planning process; and
3.	streamlining the regulations (Wijkhuijs 2003, Wijkhuijs and Glasbergen 2004).

The findings on these points are elaborated below.

2.1.1 Goals are relaxed

Most local governments embarked on the preparation of noise-abatement plans with great enthusiasm. Of the seven local authorities, however, only two came up with a policy for their entire territory. The others decided to focus on a specific sub-area and to gain some experience at that scale first; only three plans had been formally approved by 2003. Planners had to make concessions in many other respects, too. Taking both a broader definition of the problem (now embracing the quality of the living environment) and a cumulative approach led to difficulties. Several local governments conducted surveys on liveability, but this effort was of little practical use. Lacking a systematic model, they could not decide how to use the data when setting their goals.

Some objections were raised about fundamental issues. It was argued that an integrated approach to liveability should not be based on the specifics of a noise-abatement policy. It was also argued that the cumulative approach needs a quantifiable standard, which was not available. In light of the objections, all seven pilot projects opted for a focus on noise issues only. Even so, though this sectoral approach was still area-based, there was wide divergence in the way areas were circumscribed. In some cases, areas were defined in terms of the planned spatial functions; in other cases, the criterion was the current level of noise.

It is striking that current noise levels were taken as the baseline values in almost all plans. In residential areas, the focus was initially on the preferable maximum noise level. If this diverged from the current level, the goals were revised. In this way, the status quo was elevated to the status of goal. Hardly any of the plans identified problem areas, and none of them set reduction targets because of the expected cost of abatement.

Regarding to the areas to be developed, the goals for noise levels as well as the dispensation rules were formulated such that these measures would not obstruct planned development. If implementation problems were foreseen, the norms were generally relaxed, thereby using the surrounding space for noise emissions.

Three of the plans did not indicate how to achieve the goals. As a consequence, the new approach does not create a stronger noise policy but instead legitimizes a stalemate. In two cases, even the nationwide upper limit of 70 dB was adopted as the local goal for certain areas.

2.1.2 Participation is restricted

Addressing noise problems means finding the best possible balance of social, economic, and environmental interests. Accordingly, various stakeholders are involved in the process in one way or another. For this reason, the central government recommended an interactive approach to noise-abatement plans, including various government agencies but also inviting local citizens, market participants, and NGOs to take part in planning early in the process.

Although most local governments invited relevant government agencies and professional organizations such as the Chamber of Commerce and healthcare providers to join in, they were reluctant to include local citizens and their civil organizations. The low level of citizen participation can be explained by the innovative character of the planning process. The local administration was accustomed to defining noise problems in terms of decibels. In the pilot project, the authorities had to change from a technical approach to one based on a broader definition of the quality of the living environment. If that were not enough of a challenge, local planners had to determine the cumulative effects of various sources of noise and redefine an area-specific approach.

All in all, this bred technical uncertainty. Feeling vulnerable, the planners thought they needed more agency experience first. They also felt that the policy space, or scope for change, was limited, because the main sources of noise pollution, such as railways and highways, fall outside the bounds of local government control. Furthermore, it is difficult to change the noise levels deemed acceptable by industries. Finally, the local authorities realized that they did not have much to offer their local residents, in view of the high cost of abatement measures.

2.1.3 Regulations are complex

Besides improving the noise-abatement policies, the new approach was expected to make the regulations less complex. However, there is a striking degree of local variation and complexity in the rules set forth in the plans. In general, detailed conditions are formulated when the plans for new building projects allow higher levels of noise. In most cases, the new regulations are as detailed as before. While this effect is fairly common in decentralized policies, it runs counter to the goal of greater transparency. As a consequence, developers and builders will be confronted with more variety in norms and rules instead of less.

2.2 Evaluation of the results of case study 1

The analyses were focused on the innovators at the local level. These actors, who have an open mind about new forms of regulation, hardly make use of the opportunities provided by reflexivity. Instead of a simpler rule system, they end up with more variation and complexity. In most cases, lower noise levels are not foreseen. Local governments are reluctant to start an interactive process because they fear that the stakeholders will make unwarranted claims. These poor results are partly due to technical and political uncertainty.

Technical uncertainty derives from the absence of a quantifiable standard to assess the local situation. Current standards relate the dose of noise to adverse health effects. In their attempts to determine the cumulative effects of noise nuisance, the local authorities were exploring a new standard, incorporating the way noise is experienced by the local population. The premise is that different sources create different degrees of annoyance. But since no such standard has been set, the local authorities still treat the diverse sources of noise in a traditional way: separately and in terms of their health effects.

Political uncertainty derives from the lack of progress in decision-making by the central government. After the bill was published, national decision-making on noise abatement got bogged down. Because of changes in the administration, the bill never passed the obligatory advisory bodies. This left local administrations in limbo, not knowing how their plans would fit into the prospective national legal framework. Two reasons directly related to the theory of reflexive law seem to be even more important factors in explaining the poor results of the noise-abatement pilots.

The first concerns the boundaries of the system that is the object of reflexive regulation. The main sources of noise lie outside the local system. As some of the regulation also takes place outside the local system, the local authorities have less control over the policy field. While the noise impacts of societal developments may be manifest at the local level, they are often the object of policy at the local, regional, national, and even international level. Even then, government as a whole may not have a firm grip on the underlying causes of noise pollution.

This raises the issue of how to define a system that is susceptible to reflexive regulation. In reflexivity theory, subsystems are dealt with as islands that can be isolated from their surroundings. But do such islands exist in modern societies? Most societal subsystems cannot control their complex environment anymore; their decisions are somehow related to, and dependent on, decisions in other subsystems.

The second limitation of reflexivity relates to the values internal to the system. In our pilot projects, environmental concerns had to be weighed against social and economic concerns. Stringent noise-abatement goals would restrict spatial development at the local level. As local planners felt that new housing and industrial areas were urgently needed, the social and environmental goals were readily dropped.

Participatory policy-making, bringing in the values and negotiation power of citizens, could have created a more balanced approach. But as mentioned above, the cases were pilot projects, and full participation of all relevant stakeholders was not implemented in practice. The lesson to be learned is that reflexivity requires a rich value system. If a system is dominated by one set of values, lacking countervailing ones, the most likely outcome will be ‘business as usual’ instead of progressive environmental change. This is precisely what happened in the pilot projects.

3 Case study 2: Reflexivity in urban redevelopment

Noise-abatement policy has not changed, despite 10 years of debate and 4 years of experimentation. Altogether, these efforts have produced great uncertainty at the local level. Noise is still regarded as one of the country's chief environmental problems. In 2003, the bill was repealed because it could not deliver what it had promised. Hopefully, another approach will create better opportunities. On that note, let us now turn to our next case study of reflexive law—the City and Environment Law.

The underlying policy problem we focus on here is the dilemma of the compact city. Compact cities limit the proliferation of environmental problems; they preserve rural areas and conserve nature, save on energy, and allow an efficient public transport system. But a compact city also produces environmental problems: more noise, external safety risks, soil pollution, and worse air quality (de Roo 2000; Waals and Glasbergen 2002: 140 – 141). Older cities, seeking to maintain their compact footprint, often face problems when they try to redevelop run-down sections of the inner city and turn them into residential areas. National environmental norms in particular, by restricting the locations available for spatial development, seem to contravene the compact city concept (see Figure 1).

Figure 1. Stylized restricted planning space by environmental norms.

This stylized diagram illustrates the contradiction between the uniform quality norms of environmental policy at the national level and the more heterogeneous quality concepts of spatial planning at the local level. Environmental policy frequently sets quantified targets, specified as mandatory distances or buffer zones. Spatial policy relies on less quantifiable concepts, such as the efficient use of space, user-friendliness, or future development perspectives. In urban-planning practice, this often leads to a stalemate in solving environmental problems.

The question then arises whether it is possible to develop an alternative planning approach. Again, an answer is found in the introduction of reflexivity—in this case, the introduction of a self-regulatory mechanism into the planning system. As a starting point to the problem of urban redevelopment was redefined:

1.	Instead of striving to meet predetermined environmental norms, the broader concept of improving the quality of life should be the focal point.
2.	Planning should be a ‘collaborative place-making process’ (see also Healy 2000) in which environmental experts and spatial planners cooperate at an early stage and all relevant stakeholders are involved.

3.1 Experimental city and environment law

These ideas formed a platform for the introduction of the Experimental City and Environment Law. By experimental, the legislators mean that the law pertained to a specific time period (1 January 1999 – 31 December 2003) and was only applicable in specific cases (25 city-planning projects). The term experimental also relates to the substance of the law. A new methodology was introduced, which allowed for deviations from existing environmental norms if specific procedures and conditions were fulfilled.

The procedures were set forth in a mandatory three-step approach. In the first step, attempts should be made to solve the planning problem at the source. For example, to reduce noise or air pollution, one option is to lower the speed limit on a highway that cuts through a residential area; another is to relocate industries, thereby removing the external safety problems. The second step is to search for tailor-made solutions. In practice, this phase challenged the planners' creative capacity. An example was the plan to use polluted sludge and waste from a garbage dump to build a noise barrier between an industrial estate and a residential area. If the problems could not be solved in these two steps, the planner could proceed to the third step: deviation from environmental norms and regulations.

The conditions for taking the third step were formulated as rules for compensation of exceeded legal norms for noise, soil, and odour pollution, as well as external safety. If the environmental norms cannot be met in a specific environmental compartment, opportunities for compensation should be sought first in that same compartment. For example, if outdoor noise levels exceed the standards, compensation might entail installing better insulation inside the house. Or if the noise nuisance is caused by traffic, it might be compensated by measures to reduce the noise emitted by industry.

If such compensatory measures cannot be achieved, the norms exceeded in one environmental compartment might be compensated by ensuring low levels in another. For example, a high level of noise could be compensated by lower odour levels. As a last resort, the rules allow compensation in other compartments that influence the quality of life. For example, a legally unacceptable level of air pollution caused by traffic could be compensated by making the neighbourhood safer.

As an overall prerequisite, a balanced improvement in the quality of life should be attained. The measures contributing to this ‘optimal quality’ should be made explicit. Furthermore, the whole planning process should be open, involving all relevant stakeholders. When compensation is requested, medical-environmental experts must conduct a health-effect screening of the projects to document how the new situation will affect public health. The final decision on taking compensatory measures should be made by the city council; its decision is then subject to approval by the central government, i.e. the Ministry of the Environment.

With this complicated set of rules, as formulated in the Experimental Law, the cities were asked to identify problem-ridden sites that might benefit from an exemption. As the aim was to gain experience with this new planning practice, the experiment leaders selected pilot projects representing various spatial categories known to be problem-laden: railway station areas (4), harbour areas (4), inner-city areas (5), inner-city business parks (6), and post-war residential areas (6). The selected locations were subject to a range of environmental threats—noise pollution (all), odour (5), air pollution caused by traffic (10), water pollution (2), and external safety issues (7).

As is often the case with such locations, they are mainly low-income areas. The whole implementation process was supported by central government through a service centre, regular meetings of local project leaders, and funding for general studies. An independent evaluation committee was installed to make an inventory of experiences with the Experimental Law.

All pilot projects were in the implementation phase by the time the statutory limit for the Experimental Law was up. Interestingly, all but three cases proved solvable within the standard limits. One of the three projects needing compensation to work towards an acceptable outcome was a harbour area, where external noise nuisance was compensated by providing better insulation inside the houses and a better-than-usual quality of public space. Another one was an area along a river, where external noise nuisance was compensated by providing extra indoor insulation, orienting the dwellings towards water views, and improving the design of public space. The third project was a post-war residential area, where external noise was also compensated by installing inside insulation as well as by building more houses with a view of the nearby river.

3.2 Results of the pilot projects

This Experimental Law is a good example of successful reflexive regulation. Indeed, the central government of the Netherlands is giving its success wide exposure (VROM 2003). However, the official evaluations and many case study reports of the projects reveal more about the methodological aspects of the experiment than the mere fact that it had broken the stalemates. On that basis, the following analysis examines the three main concepts underpinning the experimental law: quality of life, compensation, and open planning processes.

3.2.1 Methodological problems with the ‘quality of life’ concept

The new objective—to attain an ‘optimal quality of life’—prompted the planners to take a fresh look at the old problem areas and consider whether the pending legislation would actually provide new opportunities to break entrenched stalemates. However, the law did not define the concept of quality of life. Although many projects started with an assessment of quality of life at the beginning of the planning process, the concept gradually dropped out of sight. In the absence of a generally accepted auditing procedure, each project initially developed its own assessment criteria. Later on, it proved that three fundamental methodological problems could not be solved.

The first was the problem of defining the components. At least three levels of criteria may be discerned. The physical criteria derived from classic indicators of environmental nuisance (such as measures of air, soil, and water pollution) may be assigned to the first level. Broader environmental criteria, e.g. availability of green space, quality of dwellings, and safety on the streets, may be allocated to the second level. The next level may consist of criteria for non-environmental quality, e.g. employment opportunities and the quality of public services. Although these criteria can be quantified, the relative value assigned to them will always be arbitrary.

The second methodological problem arises when these criteria are connected to the subjective feelings of the residents or stakeholders. Quality of life is also related to how it feels to visit an area or live there. An interesting outcome of the early attempts to define quality of life was the difference between quality as objectively measured and as actually experienced. Residents assigned greater importance to the quality of their homes and the appearance of the neighbourhood than to any of the classic environmental nuisances (RIGO 1999).

The third methodological problem concerns the term ‘optimal’. We can measure how much a situation improves or deteriorates, but we still do not know what the optimal point would be. Furthermore, the territorial scale used in the measurements influences how optimal quality is assessed. Altering the scale can generate divergent measures of quality. The same holds for financial opportunities. With more money available, one might expect better quality, and vice versa. Thus, an optimal quality of life, though very low, can be achieved when funding opportunities are scarce.

In light of these methodological problems, it is not surprising that the concept of quality of life remained vague. We end up with a process criterion: ensuring that quality of life is taken into account and all of its aspects are evaluated. In these terms, and with the restrictions mentioned, the evaluation committee was rather positive about the improvements in the quality of life attained in the projects (Evaluatiecommissie Stad & Milieu, 2004: 12 – 14). However, one must keep in mind that this is just an estimate.

3.2.2 Methodological problems with the ‘compensation’ concept

The main reason why local governments were eager to participate was to secure exemptions from legal environmental norms. However, although several compensation opportunities were created, only three projects took advantage of them and on only one topic, namely noise abatement.

During the planning process, it became clear that there was enough scope to deviate from the norms for soil pollution (VROM 2004: 71 – 72). Recently, an opportunity was created for functional soil cleaning—the option of varying the extent of the soil clean-up in accordance with the proposed spatial function. For air pollution, mainly NOx emissions from traffic, EU regulations obviously did not provide sufficient latitude. Setting the norms for odour emissions was already a municipal responsibility. As for external safety, no one dared propose any deviance from the norms after recent accidents in the Netherlands: an explosion at a fireworks factory and a big fire in a dance hall, both with many casualties.

But more fundamental problems—here too, of a methodological nature—had to be addressed. When discussing compensation, the first question that arose was ‘compensation for whom?’ Are the stakeholders the people who already live near the place or those who are going to live there? And if compensatory measures are taken outside the area where the problems occur, who will be better off? Next, there is a question of what to compensate and how much compensation would be suitable. How wide should a park be to compensate for odour? Is a waterfront view a better compensation? Will a nearby day-care centre really compensate for the noise problems?

In the end, there seem to be no objective criteria by which to put a value on compensation. Compensation can only be decided on in a negotiation process among current stakeholders. Apart from holding a health screening drive—and one must keep in mind that the norms are not randomly chosen but rather based on an assessment of the health risks—only a process and/or a justification criterion may be applied. A process criterion is used to check whether the planning process was really an open one in which all interested parties could have had a say. A justification criterion requires the final decision-makers to explain their reasons for opting for a compensation measure.

3.2.3 Methodological problems with the open planning processes

An open planning process, whereby all relevant local residents and organizations are involved at all stages of the planning process, was a precondition for taking the step towards compensation. Since no further procedures were established for organizing this process, planning took various forms—with mixed results. Some projects focused on individual citizens, others on particular interests, and yet others merged these vantage points. Creative solutions were produced in some cases; in others, participation led to delay or even severe opposition.

Some general problems of interactive planning came to light, too. Different stakeholders perceive quality of life differently. Because the concept is open to interpretation, the debate inevitably took on a strong ideological tone. As a consequence, the perceptions of quality that the participants brought into the process did not correspond to the general view of spatial quality; indeed, the participants had no incentive to take such a view (VROM 2001c: 5).

Obviously, an open planning process requires some form of stakeholder participation. Methodologically, however, there is still great uncertainty about how to organize the process such that it will stimulate creativity.

3.3 Evaluation of the results of case study 2

One conclusion suggested by this analysis is that the methodological innovations of the experimental law—articulated by its main concepts—did little to stimulate reflexivity. In view of the methodological problems, one might even question the worth of the concept of quality of life and the compensation principle. Nevertheless, a process of reflexive planning did take place. More actors than ever took part in an endeavour to get a wider perspective on the problem areas and the opportunities to redevelop them into more pleasant spaces. And they were rather successful in their attempts.

What, then, triggered this involvement in reflexive planning? One could argue that the main impetus came from the psychological effect of the law, combined with a specific organization of the planning process, as well as the extra attention devoted to the experimental projects. First, the prospect of compensation seems to be more important than the real opportunity to make use of it. The very idea of compensation stimulates a more open and creative approach to planning—that is, an approach less restricted by predetermined norms (ECORYS-Kolpron 2002: 18). Interestingly, the original assumption—that creative spatial planning is hampered by legal environmental norms—proved to be false. In almost all cases, creative solutions were found while the norms were met.

The second factor, the organization of the planning process, is clearly important. All of the projects were experiments that broke through the traditional organizational structures and cultures of the planning agencies. In traditional spatial planning processes, hierarchical ordering dictates when various disciplines should come into play and what their contribution should be. Each discipline looks at the situation from a different angle without seeing the problem area as a whole. Some look for green space, others focus on the design of the buildings, yet others consider the means of transport or the environmental effects. In the projects, this fragmentation had been replaced by an integrated and collaborative approach from the start of the project. The opinion of environmental professionals used to be sought at the end of a planning process; in these projects, they came into the process early and participated fully (Evaluatiecommissie Stad & Milieu 2003a/b). In fact, their position was strengthened when positive outcomes began to emerge.

The third factor stimulating reflexivity has to do with the experimental status of the projects. Their special status allowed the Ministry of the Environment to support the projects in many ways. A contact person at the Ministry helped resolve disputes, opened doors to other ministries, and made funding available to conduct extra studies to address specific problems. The regular contacts between the project managers also helped to create a sense of tackling a common challenge in a concerted effort. Finally, with their experimental status, the projects also drew considerable public and political attention, creating a general feeling that failure was not an option.

A remarkable conclusion of follow-up studies shows that the agencies involved in the experiments did not change their planning practices in other cases, although all stakeholders recognized the value of an integrated and collaborative approach. So far, the change in planning style, i.e. thinking in terms of quality of life and, particularly, arranging for connectivity between related planning professions at an early stage of the planning process, has proved to be only temporary (IMEconsult 2001: 2, NovioConsult 2003: 2, Evaluatiecommissie Stad & Milieu 2004: 16). This finding indicates how difficult it is to change planning practices, because of institutionalized working methods and underlying power relationships, outside the scope of a temporary experiment. It also shows the need to take into account the fixed organizational characteristics of the context within which a kind of reflexive planning is introduced.

4 Conclusions

In this paper, two experimental forms of decentralized reflexive regulation were studied. The first introduced a voluntary plan as the central change mechanism. The second opted for a restructuring of planning practices. Some methodological innovations are common to both, in particular the broad view on environmental quality, explicated in the concept of liveability, and the recommendation to involve local stakeholders in the policy-making process. The City and Environment Law also introduced the innovative concept of compensation and a set of rules to handle the concept. The Dutch experiments were introduced to improve the problem-solving capacities in a situation wherein detailed substantive regulation was evaluated as not productive. As we have seen, the City and Environment case showed more problem-solving capacity than the noise policy case. In the case of the noise-abatement policy, the sources of environmental problems were hardly addressed. The norms were almost completely abandoned and a participatory process not implemented. One reason is to be found in the starting position. In the City and Environment case, all stakeholders agreed that the current situation was undesirable. They perceived a common problem that connected their interests. In the case of the noise-abatement policy, this sense of urgency was absent. The starting point was not a specific problem but the general aim of prevention and improvement of the noise situation. Furthermore, the City and Environment case was well regulated. It put the emphasis on solutions at the source, first taking existing norms into account and making a participatory process mandatory to obtain compensation. A well-designed process seems to induce creative solutions, while a large amount of policy freedom, such as in the case of the noise-abatement policy, does not create this incentive. In the last instance, relaxing the environmental norms is too easy. In other words, regulations may shift from substantive rules, prescribing content, to constitutive rules, setting procedures, but the regulations still need to be forceful. A major drawback of reflexive law theory seems to be its neglect of methodological problems that need to be solved to induce reflexivity. To fully employ the potential of reflexive regulation, one should also be aware that environmental decisions are taken in the context of a web of social relationships and competing interests that need to be organized (Ost 1994). The City and Environment approach, with its combination of the concepts of liveability, compensation, and an open planning process, certainly has some potential to improve environmental problem-solving capacities. However, some conditions need to be fulfilled to fully employ the potential:

1.	Conceptual clarity on the concept of liveability: The concept of quality of life, though opening up opportunities to take a broader view on classic environmental issues, has a vaguely circumscribed meaning. Various perspectives are introduced into the discussion (related to scientific knowledge, individual experience and societal interests), as well as various fundamental principles (such as justice, equity, and equality). These perspectives and principles vary over time and from place to place (Driessen 2005: 15). One might state that defining quality of life is a matter of societal debate and political choice. However, even then some conceptual clarification and specification, defining the aspects, scales, indicators, and causality, and some frame or methodology to weigh the components, is necessary to make the concept useful as a policy tool. Without such a methodology, quality of life is a fuzzy concept, easily masking conflicts of interest, and not giving any direction to the decision-making process.
2.	A set of forceful recommendations on how to organize a participatory process: An agency that has the chance to solve problems in a reflexive way will do so only if this fits in with its prevailing, dominant values. This became very clear in the first case study, in which the environmental problem was given less priority than the economic problem, and all environmental space available was used to make urban development possible. This problem is not solved by introducing general procedural regulations, because new aspects will not get much attention, and learning will not take place. Changes are only possible if countervailing interests get a place in the planning process. A mandatory participatory planning should ensure that the processes are open, transparent, and symmetrical. They should facilitate participation and equal treatment of all relevant interests, and a clear separation of dialogues and decisions (Glasbergen and Driessen 2005).
3.	Validation by a higher-level authority: The cases showed that to solve environmental problems, a governing agency external to the system needs to put continuous and strong pressure on the local system to bring about change. The problems that were addressed in this study, though newly defined as a primary responsibility of local government, could not be solved without the participation of central government. Central government, because crucial societal issues were at stake, had to define a framework for planning and was involved in the assessment of the results. If compensation is involved, at least a quality control, with explicit justification criteria, is necessary. These criteria should cover what data should be collected and registered, regarding both the initial situation and the situation after the project has been implemented.
4.	A set of recommendations on the collaboration of planning agencies: Often, there is no uniform and direct relationship between the problem to be solved and a specific agency. Many liveability problems are multi-sector and multi-level problems. Both cases showed that planning practices are very inflexible. Engrained working methods may also hamper any change. Outside the experiments, as became obvious in the second case study, planning practices continued as usual. The introduction of reflexive law requires changes in the structural relationships, including new working relationships, between agencies involved. One of the main challenges of reflexive planning seems to be the creation of the organizational conditions that harness a collective reflexivity of the problems, i.e. to induce project-oriented cooperation of all relevant planning agencies.

Even if those conditions are met, decentralized reflexive regulation is not useful in all situations. From the case studies, we derive a more fundamental flaw of reflexive law theory, namely its tendency to underestimate the context within which societal subsystems function. It does not fully recognize the positive and negative influences originating from that context. To be fully reflexive, these systems, in this case local governments, should be able to control their environment. Noise gives an excellent example. It is not incidental that noise is one of the main problems in modern cities. The creation of noise seems to be a consequence of our current city-planning regime, with its focus on dedicated transport systems and densely built-up residential areas. Current government policies do not have a firm grip on the driving forces behind the processes creating noise. Classic substantive regulation does not provide the answer. However, the position of reflexive law, with its focus on constitutive rules, is the same. All that the governing systems can do is to accommodate themselves as best as possible to the irreversible societal processes without solving the problems themselves. In that context, compensation can be seen as a solution only if the original concept of environmental quality, in terms of zoning restrictions, is redefined in terms of the new broader concept of quality of life. The leeway created in this way must be seen as a symbolic way out.

5 Epilogue

In the mean time, the Dutch government has structurally anchored the City and Environment approach in its environmental and spatial policies. A new law, the City and Environment Act, has come into force by the end of 2005. With this law, Dutch government hopes to achieve an economic and efficient use of space and the best possible quality of the living environment. Both urban and rural areas can benefit from the new law. By allowing conditional deviation from environmental standards for noise, air, soil, odour, and ammonia, the law enables municipalities to implement projects in areas where environmental regulations would normally stand in the way. The City and Environment Act allows local authorities to deviate from national environmental regulations under the following conditions:

•	They need to achieve an economic and efficient use of space and the best possible quality of living environment.
•	They need the permission of the province.
•	They must provide compensation in the form of other measures to improve the environment, e.g. by reducing noise/odour nuisance or air pollution, or refurbishing public spaces.
•	They must have the support of local residents for example by an open planning process.
•	They have to ask advise from the municipal health authorities on the public health effects.
•	They may not deviate from European legislation.

The new Act is an improvement in the sense that it introduces a stricter definition of the aspects of quality of life; only classical environmental issues are involved. However, the concept remains vague. It is up to municipalities and all other stakeholders to decide what makes for a good quality of the living environment in a given location and situation. Consensus on methodologies has not been realized. Only in the case of deviation from existing norms is a participatory planning process mandatory to guarantee that any relaxation of the environmental regulations enjoys consensus support: the municipality must discuss any consequences of this step, and the measures to be taken to compensate for any adverse impacts, with the residents directly involved, those living nearby, and other stakeholders. Again, uniform methodologies and decision criteria are not available. The same holds for the justification criteria. The internal organization of local planning hardly received any attention in the preparation process of the new Act. The conclusion may be that the conditions formulated are still relevant, both as additions to reflexive law theory and as aspects to focus on in the further process of reflexive environmental regulation.

Acknowledgements

This paper was mainly written while the author worked as a visiting Professor at the RMIT, Melbourne, Australia. He would like to thank John T. Jackson and David Mercer for their useful comments.