Land use and flood protection: contrasting approaches and outcomes in France and in England and Wales

Applied Geography 25 (2005) 1–27 www.elsevier.com/locate/apgeog

Land use and flood protection: contrasting approaches and outcomes in France and in England and Wales Nathalie Pottiera, Edmund Penning-Rowsellb,*, Sylvia Tunstallb, Gilles Hubertc a

University of Versailles St Quentin en Yvelines, UFR SSH-47 Bd. Vauban 78047 Guyancourt-France Research Centre “ge´ographie des milieux anthropise´s”, UMR 8141 du CNRS, France b Flood Hazard Research Centre, School of Health and Social Sciences, Middlesex University, Queensway, Enfield, Middlesex EN3 4SF, UK c University of Cergy-Pontoise, 33 Bd. du Port 95011 Cergy-Pontoise cedex-France Research Centre MRTE, France

Abstract Different approaches to the control of floodplain ‘encroachment’ exist in France and in England and Wales. In France, a ‘coercive’ approach emphasises strong central government intervention within a system of designated risk zones for all natural hazards. In England and Wales, a more ‘cooperative’ approach prevails, with the dominant power being with democratically elected local authorities. Ideas and policies are converging, however, as both local flexibility and national direction are shown to have implementation weaknesses as, in both countries, the development pressures on floodplains continue to grow. q 2004 Elsevier Ltd. All rights reserved. Keywords: Flood risk; Flood mitigation; Land use policy; England and Wales; France

Introduction This paper compares the development of floodplain policies in two adjacent countries. We set this comparison in a context of the political systems, the cultural backgrounds and some aspects of the geography of the two countries, building on a decade of

* Corresponding author. 0143-6228/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apgeog.2004.11.003

2

N. Pottier et al. / Applied Geography 25 (2005) 1–27

collaborative flood hazard research (Penning-Rowsell & Fordham, 1994; Pottier, 1996, chap. 12, 1998). The floodplain policies we examine have similar aims in the two countries (e.g. Bourrelier, Deneufbourg, & De Vanssay, 2000; Environment Agency, 2001b). They seek both to minimise the adverse impacts that floods generally bring, in terms of flood damage and disruption, and to maximise the productive use and conservation values of the areas liable to flooding (the floodplain) and those generating flood runoff—collectively termed the flood risk zones. These aims are difficult to achieve and the way that their twin objectives are addressed is different in the different countries. No arrangements are perfect, but future policy makers should be able to learn from our comparisons.

National culture, ‘place’ and flood risk management The nature of flood risk management in different countries revolves around their legislative and administrative systems, their cultural contexts, and the types of floods that they experience. In this respect, France and England and Wales obviously have significant similarities. They are both members of the EU, and they share many historical experiences; both have occupied each other and fought wars together; both have a common Roman lineage. They share a place dominated by a northern Atlantic geography and climate, and by a western democracy-based capitalist economy. Both have a history over the last 50 years of substantial investment in flood defence, and of continuing flooding during this period leading to some serious loss of life (MEDD, 2004; Tunstall, Johnson, & Penning-Rowsell, 2004). But there are also significant differences, related to many factors but in particular to the French Revolution and its aftermath (Table 1). In France, politically, there is the wellknown emphasis on liberty, equality and fraternity. In the UK, the stress is on freedom, democracy and the rule of law: some subtle differences here. In terms of the drivers of Table 1 The contrasting contexts of anti-encroachment policies in France and in England and Wales France

England and Wales

Legislative systems and legal arrangements

Laws more fully codified. Strong mandatory component; legal base in the Napoleonic Code

Administrative arrangements Political culture/rhetoric

More centralised

Common law base, with strong case law component. Many legislative elements left open, for later Ministerial interpretation More devolved. Separation of judiciary, legislature and executive Freedom, democracy and the rule of law

Types of key flood problems and threats

Liberty, equality and fraternity philosophy carried forward from the French Revolution Mediterranean summer thunderstorm events; winter floods on the major rivers

Winter floods on the major rivers; coastal flood threat from North Sea storm surges

N. Pottier et al. / Applied Geography 25 (2005) 1–27

3

current flood risk, the south of France’s Mediterranean climatic region embraces rivers with a high ratio of extreme to annual floods, and rapid rates of runoff. The Alpine fringe also brings floods driven by snowmelt. England and Wales has a serious coastal flood threat, driven by strong North Sea storm surges and declining land levels following glacial retreat. The largest French rivers are longer—and their floodplains are wider—than any in the UK, including the cross-national Rhine, the Seine, the Loire, and the Rhone/Saone system. But France’s population density is less than half that of the UK and the French urban form includes more ancient pre-industrial features, with their high town centre densities, leaving rural population densities as some of the lowest in north-west Europe. The countries have different legal traditions, with the French Napoleonic Code contrasting with the UK’s common-law base. All these parameters influence the nature of ‘place’ in the two countries, and therefore the background to the efforts of individuals and agencies to manage the risks that floods can bring and, in particular, their attempts to control the build-up of flood damage potential owing to floodplain ‘encroachment’.

The issue of floodplain ‘encroachment’ To minimise flood damage and disruption in any country (Penning-Rowsell & Fordham, 1994) means preventing land uses with high damage potential—predominantly ‘urban’ uses—from ‘invading’ the floodplain, a process often termed ‘encroachment’ (Circulaire interministe´rielle, 1994; Parker, 1995). Nature conservation objectives will also be served thereby, because the ecological richness of floodplains can therefore remain relatively undisturbed. However, a completely anti-encroachment policy approach would not be wise. First, it should be moderated to match the degree of risk. ‘Encroachment’ may not be so undesirable at the fringes of the floodplain as in the high-hazard areas, but antiencroachment policies may need extra emphasis in regions where dangerous flash floods predominate. Secondly, the strictly economic argument (Green, 1998, 2002; Green, Parker, & Tunstall, 2000) is that flood risk management should aim to maximise the performance of the whole catchment and not to minimise flood losses. This may, in some cases, be best be achieved by developing floodplains, since improved efficiency of catchment use can outweigh both increasing average annual flood losses and higher flood alleviation costs (Green, 1998, 2002). But this ‘anti-sterilisation’ argument has not found favour with government, perhaps because flood damage and distress create political imperatives towards minimising floodplain use rather than seeking an economic optimum. Antiencroachment policies therefore generally still dominate in both countries. A number of factors are nevertheless liable to promote floodplain encroachment. These include a general ignorance of the risks and the consequences of floods, or a willingness by floodplain users to gamble—in the context of the infrequency of serious floods—that they themselves will not suffer losses during their use of the flood risk areas. There is also

4

N. Pottier et al. / Applied Geography 25 (2005) 1–27

the hope that if there are serious losses, then these will be ameliorated for the individual floodplain user through state compensation, insurance or some other mechanism. Our research (Fordham, 1992; Green, Tunstall, & Fordham, 1991; Pottier, 1998) suggests that ignorance per se is not the prime factor, as is often supposed, but that individuals and organisations often overlook the floods of the distant past—say, more than 10 years ago. They erroneously conclude that these were ‘freak’ events in an unsophisticated and pre-modern era that will not recur, at least in their lifetime. If floods do occur, they are often interpreted as the result of some human mismanagement; if floods have not occurred, this is often seen as due to the unsurprising consequences of improved river management (Green et al., 1991). Governments may discourage encroachment, but local developers and community leaders will often have different interests; floodplain development brings rewards in the form of private profit for individuals through increased land values and as higher levels of property-based community taxes for local government, derived from increased employment, regeneration and other local economic change. In this policy context the central problem is therefore how to control encroachment so as to prevent serious increases in flood damage potential. If this can be achieved, governments will not need to invest their scarce resources to protect retrospectively those unwelcome new floodplain uses, as has often been the case in the past (Parker, 1995; White, 1945). The traditional solution to this problem has been to try to use land use planning or building regulation systems—or both—to prevent the conversion of areas liable to flooding to urban land uses. However, this solution is being attempted in very different ways in the different countries that we analyse here, depending—we suggest—on their national culture and their geography, although we naturally cannot explore all aspects of these factors here.

Seeking solutions to encroachment problems: contrasting key characteristics The history of anti-encroachment policies in England and Wales has been detailed elsewhere (e.g. Parker, 1995), recording their origins in and before 1947. In contrast, French floodplain policy has only developed actively since the middle of the 1980s (Commissariat Ge´ne´ral du Plan, 1997), although it has rudimentary 19th century roots. Legally, France’s national policy for controlling development in flood prone areas had long been buried in scattered legal provisions. Up to 1982, only two specific tools were extant: the 1935 Submersible Surface Plan (PSS), with a hydraulic emphasis to facilitate the disposal of flood flows, and the Risk Perimeter system derived from the 1955 Planning Code. This system was specifically designed to protect people and properties by delimiting potential risks (Gendreau, 1998). The real flood prevention policy started with the 1982 Compensation Law for victims of natural disasters. The state undertook to develop Exposure to Predictable Natural Hazards Plans (PER, commencing 1984) which combined risk zoning and associated prevention measures with land use planning and flood insurance. In a 1994 Circular, central government then asked its representatives in Departments (Prefects) to strengthen the floodplain policy. The latest law of 1995 for the Protection of the Environment has

N. Pottier et al. / Applied Geography 25 (2005) 1–27

5

reinforced these arrangements and simplified the legal system by substituting a single regulatory tool for the pre-existing procedures: the Prevention of Predictable Natural Hazards Plan (PPR), which retains many of the PER aims and principles. Further catastrophic events1 have led to a further legislative review, and the Senate adopted a new law on ‘Hazards, technological risks and damage compensation’ in May 2003. The aim is to enhance levels of compensation and encourage greater public involvement in risk reduction decisions.2 Several interrelated components now appear to underlie the thinking that the formal arrangements currently embody in both France and England and Wales (Table 2). These concern the way that the land use planning system is used, the locus of power in this respect, the role of central government, the relation with other aspects of water management, and the geographical units in play and their influence on policy development. We look below at both systems, foregrounding the French arrangements. First, France has devised a specific planning system to control the existing and future use of land and buildings in areas liable to natural hazards (significantly, not only floods). This PER system (1984–1995), and the PPR developments since 1995, is a centrally directed set of arrangements applied at the local level of the Communes (the lowest level of government). Its measures are aimed at taking risk into account in the town planning documents developed by local government. Central government requires the preparation of maps delimiting hazard zones—for floods, avalanches, earthquakes, etc.—and the imposition of corresponding land use zoning and control measures (Hubert & Reliant, 2003). Flood risk is delimited on the zoning maps by showing both floodplain storage areas (champs d’expansion des crues) and developed areas (Table 3). Both the maps and the measures are formalised, and included in a single legal planning document called the Predictable Natural Hazards Plan—PPR (MATE, 1997, 1999, 2002). In contrast, it is the normal land use planning system, not the hazard-specific arrangements as in France, that underpin all attempts in England and Wales to control floodplain encroachment. Regional planning bodies and local planning authorities of various kinds (district, metropolitan and unitary Councils), serving between 50,000 and several hundred thousand people, have the responsibility for land use planning. There are no special arrangements for the designation and ‘protection’ of flood risk zones that are akin to the PER/PPR system in France or comparable, for example, to the system of designating Conservation Areas for historic towns in England and Wales for safeguarding their ancient buildings, despite the long tradition of zoning within spatial planning in the UK. 1 For example, the explosion at a chemical factory in September 2001 (Toulouse), successive sustained flooding (winter 2000–2001 in Brittany; spring 2001 in the Somme; September 2002 in southeast departments (principally Gard)). The Gard flood disaster caused 24 deaths, created 50,000 disaster victims and affected 395 communes; damage costs were estimated at V 1,000,000,000, the largest total since the 1982 Compensation Law. 2 The law provides for enhanced damage compensation and for better public information (e.g. compulsory mention of risks when selling or renting houses, a compulsory public information meeting organised by the Mayor at least biennially, and the compulsory indication of flood marks on public buildings). The project law also plans to create ‘departmental commissions and natural hazards prevention schemes’ (‘commissions de´partementales et sche´mas de pre´vention des risques naturels majeurs’) to advise on local plans for reducing risk exposure. These commissions are to be headed by the Prefect and include all local stakeholders with, amongst others, disaster victims, insurance representatives, solicitors, and the local media.

6

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Table 2 Key characteristics and components underlying anti-encroachment policies in France and in England and Wales France

England and Wales

A specific planning system to control the use of land and buildings in areas liable to any natural hazard A ‘coercive’ approach is inherent within the strong national focus implemented by central government Non-structural flood control measures explicitly dominate over structural flood control

The normal land use planning system is the main instrument for controlling floodplain land use Local interests dominate in decision-making over land use and building decisions Central government has only an advisory role rather than using compulsion, but its grant aid arrangements favour structural schemes Land use planning dominates over water planning including flood risk management

A ‘risk catchment’ approach is taken, instead of focusing on administrative areas

The reason for this arrangement lies in the decentralised land use planning system there—focused on democratically elected local authorities—that was first properly codified in the 1947 Town and Country Planning Act (coinciding, by chance, with major floods in the winter of that year). This British land use planning system, or spatial planning system as it is now called, is intended to encompass any relevant issue, including hazards such as floods or indeed any other factor relevant to decisions about a region and a local authority’s area and its land use. It does this through the system comprising forward planning through development plans and development control decisions relating to specific applications and sites. Indeed, it is one of the system’s supposed strengths that it should be able to react to any new situation or evolving local circumstances rather that having to adopt fixed solutions.3 Recent legislation, the Planning and Compulsory Purchase Act 2004, has strengthened the role and importance of regional planning but has not changed the essential democratic base. Secondly, the French system involves a ‘coercive’ approach (Handmer, 1996; Johnson & Handmer, 2004; May & Handmer, 1992) within the strong national focus implemented by central government. The system of land use and building regulation in the hazard zones is initiated by the MEDD (Ministry for Ecology and Sustainable Development—ex MATE)4 and implemented by its decentralised services provided by Departments (the 2nd level of government). This is done under the authority of the Prefects of Departments and these, crucially, are directly nominated representatives of the central government operating locally. That regulation must be approved and applied by the local authorities at the commune scale headed by the Mayor (Maire). The relevant circulars and decrees are directives (not advice) to those communes that are chosen by central government to be subject to this system (i.e. those deemed to have significant hazard exposure, whether its Mayor has requested the arrangement or not). 3 A major review of the planning system in England and Wales was initiated by central government in 2001 (National Assembly for Wales, 2002; Office of the Deputy Prime Minister, 2001, 2002). Changes were embodied in the Planning and Compulsory Purchase Act 2004. The underlying principles described in this paper have not changed radically as a result, although there is a greater emphasis on strategic planning at the regional and thence the catchment or coastal cell level and on public involvement in planning. 4 Following the 2002 French presidential election the MATE (Ministry for Planning and Development and for the Environment) became the MEDD (Ministry for Ecology and Sustainable Development).

N. Pottier et al. / Applied Geography 25 (2005) 1–27

7

Table 3 Planning response to sequential characterisation of flood risk in France (adapted from Methodological Guide PPR, 1997 and from Methodological Guide PPR for floods, 1999) Regulated flood zone

Appropriate planning response—PPR regulation

White zone of the PPR [Insignificant (ne´gligeable) or low risk areas]

No constraints

Depth of flooding!0.50 cm. Low to medium flow velocities

Some alternative urban flood control techniques can be implemented to limit urban runoff or landslides (e.g. land drainage, retention basins)

Blue zone of the PPR [Low to medium risk areas]

Suitable for most development

Depth of flooding!0.50 cm and fast flowing

For this and higher risk-zones, the regulation principle is ‘not to urbanize exposed areas’. It must be strictly applied when buildings have an impact on natural values in floodplains or make risks worse in the whole floodplain storage areas (l’ensemble du champ d’expansion des crues), even if the risk is low

0.50 cm!Depth of flooding!1.0 m and medium flowing

If other development solutions are not possible, most development is suitable with prescribed modifications to land use and to existing and future buildings belonging to public and private owners (floodresistant construction, evacuation systems, etc.)

Urban centres in high risk areas

No directly exposed areas (e.g. plateaux and valley sides) can also be regulated when some development or building may make worse the risk downstream, in the valley. They must be included into blue or red PPR zones

Developed areas with defences

In developed areas with defences such as dikes, it is necessary to maintain a strip behind them where building is prohibited (its width depending on local conditions)

No directly exposed areas (worsening risk zones) Red zone of the PPR [High risk areas] Depth of floodingO1.0 m and fast to very fast flowing waters

Most development forbidden However, the wide bans on development (l’e´tendue des interdictions) can be discussed and varied according to local situations:

Annual probability of flooding: 1% or greater (rivers)

(a) In existing urban centres: these may be suitable for development provided there are local flood prevention, protection and safeguarding measures

Floodplain storage even with low risk and even with defences

(b) In areas where the risk may be totally controlled previous to the development

Not directly exposed areas (worsening flood risk generating zones)

(c) For land uses compatible with risk (especially agricultural or forestry activities, sports grounds and water recreation areas) (d) For developing essential transport and utilities infrastructures that must be located there (e) In flash flood areas, where bans are less easily debatable. For example, reconstructing a building destroyed by a flash flood is forbidden, as are new camping sites and other facilities open to the public. For those already existing, flood warnings/ evacuation procedures and other conditions may be required (e.g. seasonal opening) depending on the PPR

8

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Therefore, while Mayors are in charge of land use planning and development in their communes, their land use planning strategies, such as the Local Town Planning Plan (Plan Local d’Urbanisme, PLU since 20015), must comply with the state’s specific risk regulations and therefore with the PPR system directed by central government. Thus, there is a division of responsibility and some sharing of power here between the centre and the periphery that is quite deliberate. Within this coercive approach there is no financial dimension, and there are no incentives for landowners, builders or developers to keep development out of flood hazard zones. Communes and property owners in PPR regulated floodplains do not receive compensation in exchange for accepting the constraints imposed on them in applying the regulations. The system is not one of encouragement or exhortation, but one of regulation and compulsion. However, the state stresses the need for the local authorities’ agreement before it acts: representatives of central government must be in contact with the local government officials concerned from the beginning to the end of the PPR process. In England and Wales, a key characteristic, following directly from the first point above, is that there is a dominance of local interests over national concerns in decisionmaking over all floodplain use. Central government is, however, not completely without a role. Under devolved government, the Secretary of State in England and the National Assembly for Wales scrutinise the local authority development plans to gauge whether they accord with national policy as embodied in a range of advice given in the form of Planning Policy Guidance (PPG) and, since 2003, Planning Policy Statements (PPS). This advice includes specific guidance in PPG 25 to local authorities on ‘Development and Flood Risk’. The Minister (and their equivalent in Wales) also has powers to ‘call in’, modify and reject local authorities’ development plans. On appeal, the Minister can also call in, modify or revoke a development control decision if it is of more than local importance. However, these central government powers are used very sparingly, and local determination dominates. With the introduction of statutory Regional Spatial Strategies under the 2004 Act, planning at regional level is set to become more important than hitherto and local development documents will also be constrained by the requirement for local policies and plans to be in general conformity with the Regional Spatial Strategy. Nevertheless the overwhelming local influence on locally applicable decisions is set to remain. Legislation and planning guidance in the 1990s have sought to enhance consistency and efficiency by changing the relationship between local development plans and development control to give pre-eminence to the former in a ‘plan-led’ planning system, a principle maintained under the 2004 Act. Herein are tendencies towards the situation in France. However, significant floodplain decisions are still made locally at the development control stage for individual land use change initiatives. Moreover, local development plans as well as development control overwhelmingly still both have a local focus because all concerned

5

In 2001, the Law SRU, ‘Solidarity, Urban Renewal’, modified town planning documents, in particular substituting the Land Use Plan (POS—plan d’occupation des sols) for the Local town Planning Plan (PLU—plan local d’urbanisme).

N. Pottier et al. / Applied Geography 25 (2005) 1–27

9

believe that the best balance between competing uses for land can generally only be judged in the local context. The reason for this is that the democracy-based local authorities in England and Wales are deemed by central government to be the prime movers in land use affairs.6 This is because these affairs affect both the local economy and the local environment which are acknowledged and relatively undisputed fields of local self-determination. Indeed without these responsibilities, local authorities would lose much of their power and raison d’eˆtre, so they are jealously guarded. The principal focus is the optimum use of all land, determined from a local standpoint, rather than a concern for hazard reduction in designated risk zones. As the main national flood defence operating authority, the Environment Agency advises on flood risk and seeks to influence the local authority development plans and decisions, but it cannot veto them. Following from this, the third key characteristic of importance in England and Wales therefore concerns the role of central government in flood mitigation through land use control. This role is one of guidance through policy documents, circulars, PPGs and PPSs. Unlike in France there is no compulsion from central government in promoting the control of development in flood risk areas. Nevertheless, this guidance has become firmer and more detailed over the years (Department of the Environment (DoE)/Ministry of Agriculture, Fisheries and Food (MAFF), 1982, 1992; Department of the Environment, Transport, Local Government and the Regions (DTLR), 2001). Both the Department of the Environment, Food and Rural Affairs (Defra; formerly MAFF), as the central government department with responsibility for flood defence policy, and the Environment Agency, have always accepted the primacy of the local authorities in land use planning. However, they have become increasingly concerned that the net effect of the devolution of this responsibility is increased ‘encroachment’ and increasing flood damage potential, hence the firmer guidance. The third key overall French characteristic has recently been to give explicit priority to integrated non-structural flood protection measures over structural flood control and the two main non-structural measures are more rigorous land use regulation, as the central feature (see above), linked to compulsory flood insurance. Engineering strategies are not ignored, but recent flood alleviation measures have generally taken non-traditional forms (e.g. small-scale works and greater ecological consideration). The national insurance and compensation system under the law of 13 July 1982 combines insurance with ‘solidarity’ (see Gilbert & Gouy, 1998, chap. 2). Thus, it requires everyone who purchases insurance for their buildings to pay, through a common and additional premium, for the costs of potential flood losses, whether the insured are in a hazard zone or not. This is the ‘solidarity’ element (Dubois-Maury, 2002), resonating clearly with French concepts of ‘fraternity’. The fourth key element in the French system is the ‘risk catchment’ approach. Instead of focusing on administrative areas (Departments, Communes, etc.), as with the PERs, the system of PPRs has progressed towards a ‘territorial’ system for the management of risk, 6 Although Regional Planning Bodies in partnership with local authorities and with community involvement have a significant role in setting strategies at regional and sub-regional level.

10

N. Pottier et al. / Applied Geography 25 (2005) 1–27

within naturally delimited areas (e.g. river catchments) rather than political jurisdictions as in the local authority dominated situation in England and Wales. The ‘territorial management’ is designed at the geographical scale appropriate to the risk phenomenon (floods, avalanches, etc.), within a sustainable development perspective. This approach is not new. In theory, it was initiated with the implementation of PERs in 1984 (MATE, 1999), but in reality it was somewhat neglected until the 1995 law bringing in the PPRs. The concept of ‘risk catchment’ derives from the principles of ‘hydrosystem’ planning and management (which echoes the concept of ecosystem management). These have developed in France since the middle 1960s and were reinforced by the Water Law of 1992. The effect is that the system of floodplain risk zones and their land use regulations can be implemented across more than one commune—and more than one Department—whereas previously this was administratively possible but not systematically practised. What we see in the arrangements in England and Wales, in contrast, is the primacy of land use or spatial planning in the form of development plans for regions and local authority administrative areas over water planning for catchments or coastal zone cells.7 The Environment Agency, as the relevant non-departmental governmental body, has a general supervisory and regulatory duty over all matters relating to flood defence and functions relating to other aspects of the water environment such as water resources, water quality and conservation. It also has responsibility for integrated and functional planning for the water environment through various forms of catchment-based planning such as Catchment Flood Management Plans. Other bodies and groupings produce plans for other aspects of the water environment such as Shoreline and Estuary Management plans. However, all these are non-statutory plans—although local authorities are encouraged in PPG 25 to take them into account—and they are all subordinate to the land use development plans for their administrative areas. The Environment Agency as the ‘competent authority’ in England and Wales also has responsibility for River Basin Management Plans for River Basin Districts under the Water Framework Directive. However, the administrative boundaries of the bodies dealing with land use planning and all these forms of planning for the water environment are different. This makes integration of the two forms of planning more difficult and inter alia means spatial planning dominates.

Formal and informal realities But neither the French arrangements, nor the system of floodplain management in England and Wales, can be understood with reference just to the above sets of ideas as 7

The EU’s Water Framework Directive has yet to be implemented in England and Wales although the Environment Agency has been nominated as the competent authority and planning is well advanced for the development of strategies for river basin planning and management as required by the Directive.

N. Pottier et al. / Applied Geography 25 (2005) 1–27

11

embodied in formal laws, rules and regulations: nothing is that simple. Many other ‘informal’ elements are significant. First, in France there is generally an absence of an overall strategy for floodplain areas. Despite the rhetoric, as elaborated above, flood management policy is not integrated into more general spatial planning and development policy. Land use control in floodplains aims only at preventing flood damage, without considering issues such as water quality, amenity enhancement along rivers, or the preservation of natural environments, although some isolated experiments in evaluating floodplains for promoting nature conservation values show that the situation has been evolving (see Degardin & Gaide, 1999). Secondly, dramatic and extensive floods over the last 10 years (e.g. 1992, 1994–1995, 1999, 2000, 2001, and 2002) have demonstrated the absence of the sought-after link between damage compensation and flood prevention measures, as prescribed in the 1982 Compensation Law (Article 5). The insurance system essentially provides compensation for direct property damage and company losses. On floodplains regulated by a PPR (or, before, by a PER), property owners are required to modify their building designs (e.g. with flood proofing) to be in accordance with the PPR prescriptions, where the cost of these works is reasonable and therefore does not exceed 10% of the building’s value. But in reality flood damage compensation has occurred irrespective of conformity with these building and land use regulations, which were designed to reduce future flood damages where encroachment has already occurred. Finally, since the general move towards political decentralisation in France in the 1980s, central government has been encouraging dialogue between the centre (l’Administration) and local authorities at the commune level. The objective is to promote cooperation (concertation) in order to bring more communes into the risk management system. Terminology is important here: this is not negotiation (negociation) but rather something discussed as it is imposed. This necessary but also delicate step often takes time because designating a PPR sometimes calls into question certain property rights and requires compulsory purchase procedures. This concertation takes the form of a softening of the somewhat rigid system of delimiting flood hazard zones: making the methodology qualitative (i.e. judgemental) rather than quantitative (i.e. mathematically modelled). But the local officials concerned have had difficulty in changing to this approach: they want to retain the more precise methodologies, despite their slowness, rather than adopt the new methods, not least because they will be responsible for any errors. However, central government is under pressure to develop a larger numbers of PPRs. They are influenced in this by a series of recent flood and storm events and by a legal ruling that responsibility rests with the public authorities involved (the Prefects and Mayors) if risk reduction is not implemented in their localities: the measures are available for use, and those responsible are required to use them. As in France, however, both the formal arrangements and the actual practices or the evolution of events in England and Wales affect what happens on the ground. Therefore, first, in many respects the way that resources are allocated by central government to flood defence in England and Wales means that ‘cure’ dominates ‘prevention’. This is because Defra grant aid to supplement locally raised resources for flood defence has been provided

12

N. Pottier et al. / Applied Geography 25 (2005) 1–27

only for capital schemes (not annual revenue support), as dictated in the relevant legislation and interpreted by their engineers. The outcome is that for many decades engineering approaches have dominated non-structural alternatives such as catchment and floodplain land use management or source control. Expenditure on ‘solutions’ has tended to follow the creation of problems—such as by encroachment or upstream urbanisation increasing runoff—rather than it seeking their prevention: a policy of protection rather than prevention (White & Howe, 2002). However, secondly, the situation is not static and policies are constantly changing in an incremental way even without new legislation or new institutions. In the new context of sustainable development, a new Environment Agency approach involving broadly based flood risk management rather than flood defence is evolving. This, for example, is embodied in the Agency’s Thames Estuary Flood Risk Management Project to develop a long-term strategy for development and flood risk for the area east of London. This strategy will encourage or at least allow development in flood risk areas provided a pro-active strategy for flood defence investment in structural works is implemented in advance of the changes in land use that will result (Environment Agency, 2003). In 2004, central government has proposed policies embodying ‘more space for water’, again reflecting a move away from simplistic and reactive flood defence policies (Defra, 2004). In these respect performance ‘on the ground’ can either support or be at variance—one way or another—from the norms expected of formal anti-encroachment policies. For the different countries here this performance is discussed separately below, reflecting the data that is available and our somewhat different research methodologies.

Policy implementation in France: results from case studies The national situation as regards risks and regulations More than 40% of the 36,500 French communes have some concern with flooding issues and flooding is responsible for 80% of the damage attributable to French natural disasters (MATE, 1999). Until 1994, the resources allocated by central government to implement the PER regulations were insufficient to allow the system to reach the expected objectives and targets. The complexity of the regulations and this lack of resources have both contributed to severe implementation delays (Pottier, Hubert, & Reliant, 2003; Pottier, Reliant, Hubert, & Veyret, 2003). Eleven years after the enactment in 1995 of the PER, only 396 communes had an approved PER, compared with over 2000 which were known to be seriously exposed to some form of natural hazard (Commissariat Ge´ne´ral du Plan, 1997). After the 1994–1995 floods, the government increased funds to meet the target of 2000 approvals by the year 2000 (PER and PPR for all natural hazards). The target was met. By June 2003, 3798 communes had a PPR approved out of 21,700 communes with some natural hazard risk (human or otherwise). Among these, 3136 communes had a flood PPR approved (plus 679 communes with a PSS, which will be revised into a PPR) out of 16,059 communes

N. Pottier et al. / Applied Geography 25 (2005) 1–27

13

exposed to flooding. The target was then revised to 5000 approvals by the year 2005, for all natural hazards (MEDD, June 6th 2003).8 On the ground, the recent nature of the PER/PPR policy and the difficulties in implementation explain the substantial amount of post-1950 development on French floodplains: 80% of buildings there are less than 40 years old (Garry, 1993). By this measure the system of land use regulations appears at first sight to have failed in containing floodplain encroachment, although that judgement must be tempered by the fact that development appears to have stopped in high hazard areas regulated by a PER after 1984 (Hubert, 2001; Pottier, 1998). A detailed evaluation of the impacts of the French floodplain land use regulations illustrates this situation (Pottier, 1998; Pottier & Hubert, 1998). Unlike in England and Wales no systematically collected national level information is available on land use change within risk zones, except that collated by Garry (1993); primary data was therefore collected from two communes on the river Marne (near Paris) and four on the Saoˆne (near Lyon). Both are big rivers with wide floodplains, and all were chosen as representative of the application of the floodplain regulations. Our survey was designed to elucidate both facts and opinions, and the findings were compared with the national flood policy, the three main interrelated goals of which are limiting floodplain encroachment, enhancing floodplain users’ information, and protecting existing and new buildings. Policy effectiveness: land use change The general trends in land use in the study areas suggest that the successive regulations were not effective in limiting overall floodplain encroachment, whether they were the old regulations in the Saoˆne valley (PSS from 1972; PER after 1990) or the more recent in the Marne valley (PER from 1990). All six communes experienced increases in the number of buildings in the floodplain (Fig. 1), but, generally, development was at a lesser rate there than outside the hazard areas, at least over the 12-year study period (Pottier, 2000, 2002). Moreover, in most of these six communes, the proportion of new buildings permitted in the regulated floodplains decreased after a PER approval. For example, in the four Saoˆne communes between 1985 and 1995, 9.3% of new buildings were located in the floodplain before the PER approval (in 1990) as against 6.9% after. Furthermore, the development only occurred in the zones where building was permitted by the regulations. Thus, the area moderately exposed to flooding according to the regulations did see increased use density (i.e. encroachment), but the high hazard area received no further development. The reasons for the land use changes and thus the development of the floodplains were not primarily the regulations themselves, but were contextual factors. The most significant of these identified in our surveys were the unavailability of vacant land for development outside the floodplain, developments that were already approved or expected at the time of the implementation of the regulations, and local political commitment to take into account 8 Personal communication at that date from staff responsible for PPR regulations at the Ministry of the Environment (now the Ministry for Ecology and Sustainable Development).

14

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Fig. 1. Planning application records and new developments permitted in the floodplain between 1988 and 1995, from the four river Saoˆne case studies (source: Pottier, 1998).

the hazard, independently of the state regulations. In the six case studies, the results are therefore positive in that the combined influence of contextual factors and the regulations have contributed to slowing development on the floodplains and to stopping it in the high hazard areas. That can be judged a measure of success in meeting the government’s objectives. Policy effectiveness: protecting existing and new buildings In floodplains regulated by a PER, the owners—householders and companies—are required to make their sites and buildings conform with the PER provisions within 5 years of the procedure’s approval (assuming the cost does not exceed the 10% limit). In the six case study areas, only 50% of the households and 32% of the companies suggested that their property benefited from the PER measures, which they perceived had only a small and an indirect impact on them in terms of implementing for new buildings any flood proofing or similar measures. The three main variables underlying the circumstances where there was this perceived benefit were where the occupants had information on hazard exposure before moving to that location, where there had been a request for a permit for any new building, and—for households—where there was a voluntary choice to live near the river.

N. Pottier et al. / Applied Geography 25 (2005) 1–27

15

Flood experience itself was, surprisingly, less important, and information on the regulations and socio-economic variables were both unimportant. Indeed, information on the hazard potential was more decisive in the decision to implement damage prevention measures than the existence of the building permit regulations. However, what the owners actually did to protect their property from flooding was in accordance with the regulations, so the building permit prescriptions probably play a greater role than these survey results suggest. For flood proofing and other measures applied to existing buildings in the floodplain, these clearly resulted from a personal initiative taken after a flood or during unrelated building work. They were not directly related to a request for a building permit that would trigger the PER regulations. So the results here are mixed. The PER regulations appear to operate to some degree but are not necessarily seen as directly responsible for what is done, although what is done is sensible, but not as universally applied as should be the case.

Policy implementation in England and Wales: results from national data sources The national policy context for development in areas of flood risks Major flooding affecting central England at Easter 1998 (Bye & Horner, 1998) prompted a review of central government guidance on ‘Development and Flood Risk’. Further sustained flooding in autumn 2000 affecting more than 10,000 properties (Environment Agency, 2001a; Penning-Rowsell, Chatterton, Wilson, & Potter, 2002) provoked further concerns. The new guidance for England was issued as PPG 25 in July 2001. Comparable revised guidance for Wales, Technical Advice Note 15, was published in July 2004. PPG 25 recognises that there are uncertainties in flood estimation, such as owing to climate change, and that the management of development and flood risk through the land use planning system should therefore be based on the precautionary principle and judgements rather than fixed formulae, as is essentially the case in France. PPG 25 sets out a new riskbased sequential approach which differentiates between floodplain areas on the basis of the level of flood risk there, whether or not the areas are developed or defended, and on the standard of existing defences and the nature of the proposed development (Table 4). The Guidance recommends that planning authorities should give priority to allocating or permitting sites for development in the areas with no risk or low risk. It requires that ‘appropriate’ flood risk assessments should be carried out for development proposals and plans and also advises that the run-off implications of developments should be assessed for all zones and, where practicable, that this runoff should be controlled as near to its source as possible, through sustainable drainage systems. The Environment Agency is encouraging local authorities to undertake Strategic Flood Risk Assessments for their areas prior to the revision of their plans and the allocation of development sites. This new guidance signals a heightened determination and a more interventionist approach on the part of central government in this field. Furthermore, central government initiated a review of the effectiveness of its guidance for England in 2004, and is checking on the delivery of its policy objectives, through monitoring the MAFF/Defra High Level

16

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Table 4 Planning response to sequential characterisation of flood risk in England (adapted from PPG 25) Flood zone

Appropriate planning response

1. Little or no risk. Annual probability of flooding: River, tidal and coastal!0.1%

No constraints due to river, tidal or coastal flooding

2. Low to medium risk. Annual probability of flooding: River 0.1–1.0%; tidal and coastal 0.1–0.5%

Suitable for most development For this and higher risk-zones, a Flood Risk Assessment (FRA) should be provided appropriate to the scale and nature of the development with applications or land allocation in plans. Floodresistant construction and flood warnings/evacuation procedures may be required depending on the FRA These and higher-risk zones are generally not suitable for essential civil infrastructure (e.g. hospitals, fire stations, emergency depots). Where such infrastructure has to be or is already located in these areas, access must be guaranteed and they must be capable of remaining operational in emergencies due to extreme flooding

3. High risk. Annual probability of flooding, with defences where they exist: River 1.0% or greater; tidal and coastal 0.5% or greater

(a) Developed areas: may be suitable for residential, commercial and industrial development provided the appropriate minimum standard of flood defence (including warning and evacuation procedures) can be maintained over the life of the development Preference should be given to areas already defended to that standard Authorities should seek to avoid areas with significant potential for coastal managed realignment or washland creation (b) Undeveloped and sparsely developed areas: generally not suitable for residential, commercial and industrial development (apart from some recreational, agricultural and essential transport and infrastructure uses where a lower risk location is not available). Caravan and camping sites should generally not be located in these areas. Where development is, exceptionally, permitted it should have the appropriate minimum standard of flood defence and should not impede flows or reduce floodplain storage (c) Functional floodplains: built development should be wholly exceptional and limited to essential transport and utilities infrastructure that is required to be located there. Such infrastructure should be designed and constructed to remain operational even at times of flood and to result in no net loss of floodplain storage, not to impede flood flows, and not to increase flood risk elsewhere. May be suitable for some recreation, amenity and conservation uses with adequate warning and evacuation procedures. A presumption against the provision of camping and caravan sites in these areas

Target No. 12 (MAFF, 1999; Department for the Environment, Food and Rural Affairs and Local Government Association, 2003). The launch of PPG 25 thereby carries a ‘threat’ of further action in that there is a possibility that central government interventions will be further strengthened in the future.

N. Pottier et al. / Applied Geography 25 (2005) 1–27

17

As in France, with the flood extent modelling for the PPR system, a major initiative operating in parallel to PPG 25 has been the production of much more comprehensive and accessible flood maps, to allow the land use planning and other authorities to have the information on which to make sensible decisions. The Environment Agency and its predecessors have long been obliged by law to produce surveys and maps of flood risk areas (Parker, 1981; Parker & Penning-Rowsell, 1981), but the information was not publicly accessible and varied widely in quality across the country. By 2000, the Agency had provided Indicative Floodplain Maps (IFMs) to all the local planning authorities in England and Wales, based pragmatically on the approximate extent of floods with an annual probability of occurrence of 1% for rivers and 0.5% at the coast. In December 2000, these maps were published on the internet (www.environment-agency.gov.uk). However, the maps have severe and acknowledged limitations: they have not differentiated between defended and undefended areas or taken into account possible climate change effects. In both respects they should be enhanced within 2005 or shortly thereafter, but the first maps did cover the whole country, comparing favourably with the much more sophisticated but sporadically introduced PPR maps in this respect. In addition to the IFMs, in 1996 the Agency adopted a national programme for collecting data and preparing maps for priority areas. These are flood risk areas targeted for building development, and the aim here is to generate and provide more precise information. The Agency also aims to identify for most of the country an outer flood area showing the likely extent of extreme floods (with an annual probability of down to 0.1%): outside these mapped areas, there should effectively be no flood risk and no constraints on development.9 When this is achieved the Agency will have made very significant progress towards overcoming the information deficiencies on flood risk which had earlier been a barrier to the implementation of unambiguous floodplain protection policies, and which had also limited the Agency’s ability to oppose successfully any unwise encroachment proposals. The aggregate effect: new information on the national situation in England As indicated, there is no national data source on encroachment issues in France. But, recently published national data provides evidence on both the level of pressure to develop in floodplains in England10 and on the effectiveness of the evolving policies against floodplain encroachment (Environment Agency, 2001b). For its IFMs in England the Environment Agency has mapped data for all residential, commercial and industrial planning applications (not approvals) for large-scale developments (with a value of £100,000 or more, or covering a large area) for a 6-year period from 1996–1997 to 2001–2002. Not all the planning applications could be included due to incomplete location data, and the results do not take into account existing defences 9 However, there is evidence from the autumn 2000 floods that many areas flooded were affected by pluvial flooding not directly related to a major watercourse and this will cause difficulties for development control decisions in all such areas. 10 Unfortunately, comparable data is not available for Wales.

18

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Fig. 2. Residential and non-residential development applications in the floodplain in England between 1996 and 2002 (financial years).

or other mitigation measures. They do, however, offer valuable estimates of trends (Environment Agency, 2001b, 2002). These data reveal that, on average, about 10% of all development applications in England are for floodplains, and they show growing development pressure on floodplain land in England (Fig. 2). The proportion of all applications that are for floodplains has increased from around 8% in 1996–1997 to around 13% in 2001–2002. The number of housing units proposed (not actually built) for flood risk areas has increased almost sixfold over the period. In the year 2001–2002, developers applied to build nearly 36,000 housing units in the indicative floodplain, worth about 4.5 billion Euros in construction costs (c. £3 billion). Applications to build commercial and industrial property in these floodplain areas quadrupled over the same period, and in the final year there were applications for 2700 projects with an approximate construction value of 12 billion Euros (£8 billion). The trends in land use planning: High Level Target No. 12 As evidence of central government’s increasing concern in this field, the Defra/MAFF High Level Target 12 seeks to gauge the effectiveness of policies against floodplain encroachment in England (MAFF, 1999). The Target, firstly, requires the Environment Agency, with local authorities, to report on those local authority development plans on which the Agency has commented, identifying those which do and do not contain flood risk statements or policies. Secondly, it requires reports on the Agency’s responses to planning applications, identifying cases where it sustained objections on flood risk grounds and where the final decisions—either by the local

N. Pottier et al. / Applied Geography 25 (2005) 1–27

19

Table 5 Planning applications and decisions where flood risk was an issue in Environment Agency consultations in England (source: Environment Agency, 2002, 2003; Environment Agency and Local Government Association, 2003) Planning applications and Agency objections on flood risk grounds

2001/2 No.

2002/3 No.

Planning applications on which the Agency was consulted Applications where flood risk was an issue Agency made objections on flood risk grounds

89,562 24,138 2500

90,164 20,452 4532

Decisions notified by the local authority to the Environment Agency in 2001/2 and 2002/3 where the outcome is known on the planning applications to which the Agency objected on flood risk grounds Agency objections on flood risk grounds and outcomes

Total Agency objections where a decision was notified and outcome was known Agency objections resolved through negotiations, withdrawal or conditions before a formal decision Agency sustained objections on flood risk grounds to a formal decision (A) Decision by local authorities in line with EA advice: e.g. application refused or granted subject to fully mitigating conditions Decision by local authority contrary to EA advice: application granted or granted with only partially mitigating conditions (B) Decision by local authority contrary to EA advice (% of sustained objections) (i.e. B/A)

2001/2

2002/3

No.

%

No.

%

1712

100

1952

100

954

56

905

46

758

44

1047

54

475

28

826

42

283

17

221

11

37

21

planning authority or on appeal—were in line with or contrary to the Agency’s advice.11 On development plans, it appears that most English local planning authorities adopting plans in 2000–2001, 2001–2002 and 2002–200312 had taken the Agency’s advice and information into account: almost all plans monitored contained some flood risk policies (29 out of 31 in 2000–2001; 21 out of 23 in 2001–2002; and 15 out of 16 in 2002–2003). On planning applications for specific sites and developments, the evidence is similar. The Agency responds to nearly 100,000 planning applications each year. For about one quarter of these, flood risk was an issue for particular consideration and for an increasing proportion of these, the Environment Agency made objections to the planning application on flood risk grounds (Table 5). Table 5 shows that in approximately half of the cases where the outcome on planning applications to which the Agency objected initially on flood risk grounds was known (the data are incomplete), the Agency’s objections were resolved through withdrawal of the application, negotiated changes or through the imposition of mitigating conditions before 11 The first year of monitoring covered a transitional stage during which the new central government PPG 25 guidance was implemented. The second year constitutes the first in which the new guidance was in operation over the whole monitoring period. Thus, it is still a little early to assess definitively the effectiveness of that new policy and its implementation. 12 Financial years, from April to March.

20

N. Pottier et al. / Applied Geography 25 (2005) 1–27

a decision was made. Where the Agency’s objections were maintained to the point where the local planning authority took a decision on the application, in about a third to a fifth of these instances, planning permission was granted by the local authority despite the Agency’s sustained objections. However, the number of such applications was small in both years (Table 5) as was the number of applications granted on appeal against the Agency’s advice (only 4 out of 50 in 2002–2003). Furthermore, the proportion of cases in which the Agency’s advice is ignored appears to be declining. While most of these applications that were approved against the Agency’s advice were for minor land use changes (e.g. single dwellings; garages; extensions or conservatories; or developments of several houses in lower risk areas), in both years there were some significant developments. Most of the decisions contrary to Agency advice were for residential buildings and they, surprisingly, included, in 2001–2002, a special housing scheme for the elderly. The main reasons for the Agency’s objections were the flood risk in the development area itself or the lack of an appropriate flood risk assessment (Environment Agency, 2002, 2003). Thus, in general, the Agency’s advice on development and flood risk is being accepted and followed by local planning authorities. However, this does not necessarily mean that development is being kept out of high-risk floodplain areas, particularly those that are already developed and/or defended. The huge Thames Gateway development area along the Thames Estuary in east London exemplifies this (Environment Agency, 2003). There is a growing acceptance in England and Wales that where floodplain development is allowed to go ahead, it is, in most cases, with an appropriate standard of mitigation measures, generally funded by the developer, to deal with flood risks at the site or generated elsewhere through the development. But of course these defended developments may be at greater risk in the more extreme events that may occur with greater frequency in the future with climate change, and may thus generate demand for flood warning, flood defence and emergency planning services. The ‘escalator effect’ demonstrated by Parker (1995) may therefore still be destined to continue.

Policy effectiveness: floodplain information for the public The effectiveness of any policy in the land and water management field can be judged only partly by what happens ‘on the ground’: in this case by the incidence of encroachment in flood risk areas. But this suite of policies—as others—may be effective in a different way, by informing the public of the risks that they face, such that they do not even apply for permissions for such developments. Methodologically this is problematic, because this type of effectiveness cannot be judged just by what happens in the areas affected: it is what has not happened that counts. But one of the ways that this type of effectiveness can be judged—imperfect though this is—is the degree to which the public is informed about such issues, and on this we have some data for each of our countries, albeit data that is not strictly comparable.

N. Pottier et al. / Applied Geography 25 (2005) 1–27

21

Fig. 3. Awareness of living in a flood risk area and spontaneous awareness of the Environment Agency as providing information about flooding among residents in flood risk areas in England and Wales (source: BMRB International Limited, 2001b, 2003a,b).

Public information on flood risk in England and Wales In England and Wales the Environment Agency has recognised the need to convince the public about the threats of flooding and therefore the need to use floodplains wisely. Since 1999 it has mounted annual national and local media campaigns in the autumn, to raise public awareness of flood risks and the flood warning system, particularly among those who live or work in flood risk areas and to encourage the public themselves to take preventative action. It has also monitored the success of its campaigns through systematic market research (e.g. BMRB International Limited, 2000, 2001a,b, 2003a,b). The data from these seven annual surveys (Fig. 3) show that between a half and over three quarters of those who lived in flood risk areas were aware of the risk there.13 Awareness levels may be raised further still in the future, because it is likely that those selling properties in England and Wales will be required to disclose a flood risk under new legislation now being discussed in parliament. Spontaneous awareness of the Environment Agency, a relatively new organisation formed in 1996, as the body that provides information about flooding has grown over the period of the campaigns among those in flood risk areas, although more people there still believe the responsibility to lie with the local authorities (BMRB, 2001b, 2003a,b). With much more information available through the campaigns and other publications, most of those in at-risk areas were aware of precautions that they could take in advance of flooding, although even by 2003, very few (only 8%), had actually taken any preventative action (BMRB, 2003a,b). 13

The decline in recent years is likely to be due to an expansion of the database from which the survey population was drawn to include those in areas where the Environment Agency does not provide a direct flood warning service.

22

N. Pottier et al. / Applied Geography 25 (2005) 1–27

The Agency’s recent research has not examined public awareness of, or attitudes towards, policies on development and flood risk. However, current and past research concerning both those affected by flooding and those living in flood risk areas has shown such residents generally to be supportive of some restrictions on development in flood risk areas (Tunstall & Fordham, 1994; Tunstall, Parker, Twomey, Doizy, & Sangster, 1991). Those who had experienced flooding were concerned that others should not be put at risk and that floodplain development could have exacerbated their own problems, or could do so in the future (Tapsell & Tunstall, 2001; Tapsell, Tunstall, Handmer, & Penning-Rowsell, 1999; Tunstall et al., 1991). Under the Planning and Compensation Act 2004 much greater emphasis is placed on public involvement in the development plan process. Therefore, the awareness and attitudes of residents in flood risk areas are likely to be of increasing importance as they will be in a position to put pressure on all the relevant authorities to pay more attention to flood risk issues in their decision-making. Policy effectiveness: improved floodplain information in France To be more effective, the Ministry for Ecology and Sustainable Development in France has also recognised that floodplain occupants and landowners must respect its risk management regulations, and in our six case study communes we investigated the level of knowledge of hazards and of the regulations themselves. We found that most floodplain occupants were aware of their floodplain location (75% of the households; 50% of the companies), but that the floodplain occupants’ information about the risk exposure had not increased by a statistically significant amount since the implementation of the PERs. For 64% of the households and for 44% of the companies the choice of locating in the floodplain did not result from ignorance of the hazard, since they were aware of the risk before moving into the area. Information on risk exposure was primarily passed on by the previous occupant of the building or site, or by the owner to the new tenant. For those becoming aware of the risk after moving, information came mainly from neighbours or directly from experiencing a flood event. Public authorities obviously should play a significant role concerning information about regulatory tools. Our floodplain occupants were generally aware of the existence of anti-encroachment measures, first through the communes’ municipal departments and, with equal importance, through the media. But the number aware of the existence of the PER in their commune was less than 7%, despite almost all of them moving there before its implementation, which occurred on average some 12 years later. Our data shows that the implementation of the PERs appears to have had only a weak impact on the information base of floodplain occupants, and that individuals, when responding to the hazards they face, did not do so because of the regulations, despite the compulsory nature of some of their provisions. Nevertheless, we are dealing with an evolving situation, and there has been recent progress in terms of improved public information on hazards and regulated risk areas in France. First, the Ministry for the Environment created a website dedicated to ‘major risks’ in June 1998, to give the public access to a census of communes at risk (www.prim.net). Secondly, since 1999, the Ministry for the Environment

N. Pottier et al. / Applied Geography 25 (2005) 1–27

23

has published on-line ‘lessons learnt’ reports after each flood disaster (e.g. MEDD, 2004).14 The objective is to assess causes, damage costs, and crisis management problems, and to provide recommendations for policy improvement. Thirdly, more and more Departments and Communes have developed their own hazard and PPR websites displaying the official hazard maps and their risk reduction measures (see for example www.auvergne.pref.gouv.fr).

Conclusions International policy comparison is not easy, not least because data is rarely comparable. But when we look just at our own country in a complex policy field such as floodplain encroachment we often see just the limitations as to what has been achieved, and underplay the progress that is being made (White & Howe, 2002). Comparison puts the efforts of individual countries in context and can lead to sounder conclusions. This paper has demonstrated that preventing unwise floodplain development touches fundamental issues concerning the liberty of individuals to use their land and property as they wish, even if it causes them harm, and the role of the state in preventing this from happening. This emphasises culture and politics, rather than just geography and place. Therefore, our account of attempts to prevent floodplain encroachment is redolent with issues such as fraternity in France and democracy in England and Wales, as well as the different floods from which each suffers. In this way, ‘place’ matters, because place involves both geography and political systems, and these are both at the heart of floodplain management. Taking account of place in understanding and implementing flood mitigating means, for example, that the PER/PPR system allows for Alpine avalanches, wide floodplains and the high hazard of summer thunderstorm runoff. It also means that London’s Gateway development will be allowed to encroach the Thames Estuary floodplain owing to the national interest in allowing space for London to expand there. The lower threat of loss of life in floods in England and Wales, vis-a`-vis France, means that floodplain mapping can be more pragmatic there, whereas in France the cartography is more precise but the coverage remains incomplete. But place does not over-rule the countries’ political systems and traditions, which in the end dominate policy making. Despite some methodological differences between the research in the different countries the comparison is striking. In France there is a centralised system of risk zone designation and mandatory procedures, including compulsory flood insurance, in the French tradition of central government reaching right down to the local level. In England and Wales the philosophy is one of local attention to flooding and land use within a system embodying central government guidance, not compulsion, and devolved decisions about flood alleviation measures. 14

The authors (except from G. Hubert) assisted with the ‘lessons learnt’ exercise for the flood disaster in the Gard department in September 2002 (MEDD, 2004). http://www.environnement.gouv.fr/actua/cominfos/dosdir/ DIRPPR/dosdppr.htm#risqnat.

24

N. Pottier et al. / Applied Geography 25 (2005) 1–27

The French system has an emphasis on coercion, whereas the situation in England and Wales stresses inter-institutional cooperation (Handmer, 1996; Johnson & Handmer, 2004). In the last resort one cannot judge one set of arrangements ‘better’ than another. From our case studies, the PER/PPR procedures appear to be keeping urban land uses out of the high hazard areas, but are not so successful in the lower risk areas or succeeding in their aim of promoting damage reduction for existing buildings. The trends in England and Wales generally appear to be for flood risk issues now to be appropriately addressed in planning decisions, with some notable but rare exceptions (especially in defended areas). But this has often meant development has had to be coupled with appropriate flood defences rather than being excluded from areas of risk: retrospective protection rather than proactive prevention (White & Howe, 2002). However, policies are evolving continuously. In 2003, the Environment Agency in its new Strategy for Flood Risk Management signalled a shift away from defending against flooding, with an emphasis on engineering, towards managing flood risks using a range of sustainable solutions including changing and managing land use for catchments and coastal cells (King, 2003). Policy evolves, therefore, even if formal systems appear unaltered (Defra, 2004). Public and media attitudes, faced with serious floods since 2000 in both countries, suggest that any flooding is unacceptable in modern countries. Government perceptions appear to be that that there has been poor progress in local control of risk, and as a result there is some evidence of convergence of thinking and approaches in the two countries. In France the concertation procedure is bringing some increased flexibility to what is a fairly rigid system, whereas in England and Wales the PPG 25 process is bringing more rigorous central intervention to a system where local flexibility is perceived to have created the encroachment problems that central government is left to ‘solve’. In both countries the pressures on floodplains continue to grow, driven by growing economies, a shortage of land suitable for development, and socio-demographic changes that create a requirement for more housing. There is also a growing international consensus that climate change can bring increased flood probabilities (e.g. Evans et al., 2004). Against these powerful forces, the new, fragile and relatively untested systems that have been implemented in the last decade or so to control development in areas with a risk of flooding have undoubtedly yield some successes, but they are yet to prove themselves strong enough to stem the tide.

References BMRB International Limited (2000). Flood action week 2000: Campaign evaluation October 2000. London: BMRB International Limited. BMRB International Limited (2001). Flood action 2001: Campaign evaluation October 2001. London: BMRB International Limited. BMRB International Limited (2001b). At risk 2001 report. London: BMRB International Limited. BMRB International Limited (2003). Flood awareness 2003: Campaign evaluation December 2003. London: BMRB International Limited. BMRB International Limited (2003b). At risk 2003 report. London: BMRB International Limited.

N. Pottier et al. / Applied Geography 25 (2005) 1–27

25

Bourrelier, P. H., Deneufbourg, G., & de Vanssay, B. (2000). IDNDR objectives: French technical sociological contributions. Natural Hazards Review, February, 18–26. Bye, P., & Horner, M. (1998). Easter 1998 floods: Report by the independent review team to the Board of the Environment Agency, Vols. I and II. Bristol: Environment Agency. Circulaire interministe´rielle (Ministe`re de l’Inte´rieur, Ministe`re de l’Environnement, Ministe`re de l’Equipement) du 24 janvier 1994 destine´e aux pre´fets (1994). La pre´vention des inondations et la gestion des zones inondables (6 pp.). Paris: Ministe`re de l’Inte´rieur, Ministe`re de l’Environnement, Ministe`re de l’Equipement. Commissariat Ge´ne´ral du Plan (1997). La pre´vention des risques naturels, rapport d’e´valuation. Comite´ interministe´riel de l’e´valuation des politiques publiques, Premier Ministre, la Documentation Franc¸aise, Paris, France (702 pp.). Degardin F., & Gaide P. A. (1999). Valoriser les zones inondables dans l’ame´nagement urbain. Repe`res pour une nouvelle de´marche. Dossiers Eau et ame´nagement CERTU, MATE (Ministe`re de l’Ame´nagement du Territoire et de l’Environnement) (232 pp.). Department of the Environment (DoE)/MAFF/Welsh Office (1982). Circular 17/82: Development and flood risk. London: Department of the Environment (DoE) /MAFF/ Welsh Office. Department of the Environment (DoE)/MAFF/Welsh Office (1992). Circular 30/92: Development and flood risk. London: Department of the Environment (DoE)/MAFF/Welsh Office. Department for the Environment, Food and Rural Affairs (Defra) (2004). Making space for water: A new government strategy for flood and erosion risk management in England. London: Defra. Department of the Environment, Transport, Local Government and the Regions (DTLR) (2001). Policy and planning guidance 25 (PPG 25). London: Department of the Environment, Transport, Local Government and the Regions. Dubois-Maury, J. (2002). Les risques naturels en France, entre re´glementation spatiale et solidarite´ de l’indemnisation. Annales de ge´ographie, 111e`me anne´e 627/628, 637–651. Environment Agency (2001a). Lessons learnt: Autumn 2000 floods. Bristol: Environment Agency. Environment Agency (2001b). High Level Target 12: Development and flood risk Report to Department of the Environment, Food and Rural Affairs (Defra) and DLTR, July 2001. Bristol: Environment Agency. Environment Agency (2002). High Level Target 12: Developments in areas at risk of flooding Report to Defra and the Office of the Deputy Prime Minister (ODPM), July 2002. Bristol: Environment Agency. Environment Agency (2003). Thames Gateway and flood risk management—A preliminary assessment. Bristol: Environment Agency. Environment Agency and Local Government Association (2003). High Level Target 12: Development and flood risk Joint Report to the Department of the Environment, Food and Rural Affairs (Defra) and the Office of the Deputy Prime Minister (ODPM) August 2003. Bristol: Environment Agency. Evans, E. P., Ashley, R., Hall, J., Penning-Rowsell, E. C., Saul, A., & Sayers, P. (2004). Future flooding; Scientific summary Future risks and their drivers, Vol. 1. London: Office of Science and Technology. Fordham, M.H. (1992). Choice and constraint in flood hazard mitigation: The environmental attitudes of floodplain residents and engineers. Unpublished PhD. Thesis, Middlesex University. Garry, G. (1993). Le risque d’inondation en France. Recherche d’une approche globale du risque d’inondation et de sa traduction cartographique dans une perspective de pre´vention. The`se de doctorat de Ge´ographie de l’Universite´ Panthe´on-Sorbonne-Paris I (434 pp.) plus annexes. Gendreau, N. (1998). La gestion du risque d’inondation et l’ame´nagement des cours d’eau. Annales des Ponts-etChausse´es, 87, 53–59. Gilbert, C., & Gouy, C. (1998). Flood management in France. In U. Rosenthal, & P. t’Hart (Eds.), Flood response and crisis management in Western Europe. Berlin: Springer, 15–56. Green, C. H. (1998). The economics of floodplain use. Himganga, 1(3), 4–5. Green, C. H. (2002). Making better choices: The role of economics in integrated flood management. Geneva: Global Water Partnership/World Meteorological Organisation. Green, C. H., Parker, D. J., & Tunstall, S. M. (2000). Assessment of flood control and management options. Report prepared for the World Commission on Dams. Capetown: World Commission on Dams. http://www.dams.org Green, C. H., Tunstall, S. M., & Fordham, M. (1991). The perceptions of the risks of flooding: Which risk and whose perception? Disasters, 3(15), 227–236.

26

N. Pottier et al. / Applied Geography 25 (2005) 1–27

Handmer, J. W. (1996). Policy design and local attributes for flood hazard management. Journal of Contingencies and Crisis Management, 4(4), 189–197. Hubert G. (2001). Le risque d’inondation, Evaluation globale de l’efficacite´ et des impacts de la politique re´glementaire et appropriation par la socie´te´ civile, 1er rapport d’avancement, MATE (20 pp.). Paris: MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement). Hubert, G., & Reliant, C. (2003). Cartographie re´glementaire du risque d’inondation: De´cision autoritaire ou ne´gocie´e? Annales des Ponts-et-Chausse´es, 105, 24–31. Johnson, C. L., & Handmer, J. W. (2004). Coercive and cooperative policy designs: Moving beyond the irrigation system. Irrigation and drainage, 52, 193–202. King, D. (2003). Strategy for flood risk management: A paper for the Environment Agency Board. Environment Agency: Bristol. MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement), Ministe`re de l’Equipement, des Transports et du Tourisme (1997). Plans de pre´vention des risques naturels pre´visibles (PPR), Guide ge´ne´ral. La Documentation Franc¸aise (78 pp.). Paris: MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement). http://www.ladocumentationfrancaise.fr MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement), Ministe`re de l’Equipement, des Transports et du Tourisme (1999). Plan de pre´vention des risques naturels (PPR). Risques d’inondation, guide me´thodologique. La Documentation Franc¸aise (124 pp.). Paris: MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement). MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement), Ministe`re de l’Equipement, des Transports et du Tourisme (2002). Plans de pre´vention des risques naturels (PPR). Risques d’inondation, mesures de pre´vention. La Documentation Franc¸aise (160 pp.). Paris: MATE (Ministe`re de l’Ame´nagement du territoire et de l’Environnement). May, P., & Handmer, J. W. (1992). Regulatory policy design: Cooperative versus deterrent mandates. Australian Journal of Public Administration, 51(1), 43–53. MEDD (Ministe`re de L’Ecologie et du De´velopment Durable) (2004). Crues du Gard 2002: Retour d’expe´rience. Paris: MEDD (Ministe`re de L’Ecologie et du Development Durable). Ministry of Agriculture, Fisheries and Food/Welsh Office (1999). High Level Targets for Flood and Coastal Defence and Elaboration of the Environment Agency’s Flood Defence Supervisory Duty. London: MAFF. National Assembly for Wales (2002). Planning: Delivering for Wales. Cardiff: National Assembly for Wales. Office of the Deputy Prime Minister (2001). Planning Green Paper: Delivering a fundamental change. London: Office of the Deputy Prime Minister. Office of the Deputy Prime Minister (2002). Sustainable communities: Delivering through planning. London: Office of the Deputy Prime Minister. Parker, D. J. (1981). The value of hazard zone mapping: Water Authority ‘Section 24(5) surveys’ in England and Wales. Disasters, 5(2), 120–124. Parker, D. J. (1995). Floodplain development policy in England and Wales. Applied Geography, 15(4), 341–363. Parker, D. J., & Penning-Rowsell, E. C. (1981). Specialist hazard mapping: Water authorities’ land drainage surveys. Area, 13(2), 97–103. Penning-Rowsell, E. C., Chatterton, J., Wilson, T., & Potter, E. (2002). Autumn 2000 floods in England and Wales: Assessment of national economic and financial losses. London: Middlesex University Flood Hazard Research Centre. Penning-Rowsell, E. C., & Fordham, M. F. (1994). Floods across Europe: flood hazard assessment, modelling and management. London: Middlesex University Press. Pottier, N. (1996). Land use controls. In E. Penning-Rowsell, Improving flood hazard management across Europe. Commission des Communaute´s Europe´ennes: Programme Environnement (EV5V-CT93-0296), 1–24. Pottier, N. (1998). L’utilisation des outils juridiques de pre´vention du risque d’inondation: e´valuation des effets sur l’homme et l’occupation du sol dans les plaines alluviales. The`se de doctorat de Sciences et Techniques de l’Environnement, de l’Ecole Nationale des Ponts et Chausse´es, Vol. 1 (p. 436) plus annexes. Pottier, N. (2000). Risques d’inondation, re´glementations et territoires. Hommes et Terres du Nord no. 2/2000 (pp. 93–101).

N. Pottier et al. / Applied Geography 25 (2005) 1–27

27

Pottier, N. (2002). Gestion du risque d’inondation et maıˆtrise de l’urbanisation dans le val de Saoˆne. In J. P. Bravard, J. Combier, & N. Commerc¸on (Eds.), La Saoˆne: Axe de civilisation (pp. 197–213). Presses Universitaires de Lyon, 197–213. Pottier N., & Hubert G. (1998). Evaluation de l’efficacite´ des mesures re´glementaires de pre´vention des risques d’inondation: Synthe`se des e´tudes de cas mene´es dans le Val de Saoˆne et dans la valle´e de la Marne. CERGRENE-ENPC-ENGREF. Ministe`re de l’Ame´nagement du Territoire et de l’Environnement, DPPR, lettre de commande no. 41–97 (122 pp.) plus annexes. Pottier, N., Hubert, G., & Reliant, C. (2003). Quelle efficacite´ de la pre´vention re´glementaire dans les zones inondables? Ele´ments d’e´valuation. Annales des Ponts et Chausse´es, 105, 14–23. Pottier, N., Reliant, C., Hubert, G., & Veyret, Y. (2003). Les plans de pre´vention des risques naturels a` l’e´preuve du temps: Prouesses et de´boires d’une proce´dure re´glementaire. Annales des Ponts et Chausse´es, 105, 40–48. Tapsell, S. M., & Tunstall, S. M. (2001). The health and social effects of the June 2000 flooding in the North East region. Report to the Environment Agency. London: Middlesex University Flood Hazard Research Centre. Tapsell, S. M., Tunstall, S. M., Handmer, J., & Penning-Rowsell, E. C. (1999). The health effects of the 1998 Easter flooding in Banbury and Kidlington. Report to the Environment Agency. London: Middlesex University Flood Hazard Research Centre. Tunstall, S. M., & Fordham, M. H. (1994). Thames perception and attitude survey-Datchet to Walton Bridge. London: Middlesex University Flood Hazard Research Centre. Tunstall, S. M., Johnson, C. L., & Penning-Rowsell, E. C. (2004). Flood hazard management in England and Wales: From land drainage to flood risk management. Proceedings of the world congress on natural disaster mitigation, Vigyan Bhawan, New Delhi, India, 19–22 February 2004 (pp. 447–454). Tunstall, S. M., Parker, D. J., Twomey, G., Doizy, A., & Sangster, E. (1991). The Severn floods of January and February 1990. London: Middlesex University Flood Hazard Research Centre. White, G. F. (1945). Human adjustment to floods. Department of Geography Research Paper No. 29. Chicago: University of Chicago. White, I., & Howe, J. (2002). Flooding and the role of planning in England and Wales: A critical review. Journal of Environmental Planning and Management, 45(5), 735–745. www.environment-agency.gov.uk www.prim.net