Recurrent Governance Challenges in the Implementation and Alignment of Flood Risk Management Strategies: a Review

Flood risk prevention focuses on reducing the consequences of floods through land use and spatial planning measures. These can focus on forbidding or discouraging the location of vulnerable societal functions (e.g. housing, power plants) in flood-prone areas, the establishment of building requirements or the dis-incentivising of urban development in vulnerable areas. Flood risk prevention is claimed to be the potentially most effective FRMS (Hooijer et al. 2004; Beucher 2009; Penning-Rowsell and Pardoe 2014).

However, in practice, implementation of flood risk prevention seems to be complicated. In a UK context it was found that urban development in flood-prone areas continues to be allowed (Burningham et al. 2008). In various other European countries, it was shown that parties interested in the continuation of urban development implicitly tend to believe that “water managers should enable spatial planning wherever spatial planners want them to do this” (Wiering and Immink 2006: Woltjer and Al 2007; Pardoe et al. 2011). Baubion (2015) for instance points at the presence of financial mechanisms that automatically compensate for flood losses and therefore trigger ongoing building activities in flood prone areas in Paris. In operational terms, three main barriers for achieving the implementation of flood risk prevention emerge from the literature.

A first barrier is that water managers and spatial planners generally operate through different modes of governance: water management in Europe tends to be centralised, while local governments have the dominant role in spatial planning (Wilson 2006; Storbjörk 2007).

Second, in many countries, including Italy and Germany, policy and legal frameworks for establishing risk prevention through pro-active spatial planning are limited in scope (Mysiak et al. 2013; Sapountzaki et al. 2011). In countries where such rules are in place (e.g. Plan de Prévention des Risques in France or Planning Policy Statements in England), implementation gaps have been reported. These are due to a lack of support from local planning authorities, tensions between national and local governments, a lack of coordination between stakeholders and a lack of clarity regarding hazard assessments (Beucher 20,090). In addition, the implementation of specific procedural instruments like the Dutch and Belgian Water Test – an instrument that explicitly asks spatial planners to reconsider water interests in spatial planning – has been found to be problematic (Hegger et al. 2013). Moreover, existing planning rules are often said to lack flexibility (Van den Brink et al. 2013; Harris and Penning-Rowsell 2011).

Third, terminologies and the ways of working of spatial planners differ from those of water managers. Spatial planners talk about “land-use functions” and “reconciling and combining functions” (Hartmann 2009); “land-use planning procedures”, “zoning measures”, and “environmental impact assessments” (Wiering and Immink 2006:427) while water managers’ prime focus is on “flood risks” (ibid).

Some authors however, see the obligation stemming from the EU Floods Directive to produce flood risk management plans by 2015 as a promising development, as it may lead to “spatial water governance”, a hybrid between the centralised mode of water managers and the interactive mode of spatial planners (Hartmann and Driessen 2013:7). Others point at positive experiences, gained amongst others in a Dutch context, with bringing water managers and spatial planners together by enhancing communication and learning through climate adaptation atlases or flood maps, the establishment of special learning and action alliances, ‘networks of expertise’ or interaction settings like climate ateliers (Potter et al. 2011; Van Herk et al. 2011; Goosen et al. 2014).

More emphasis on flood defence may entail improving existing structures (heightening or strengthening dikes) or applying new options such as delta dikes, unbreakable dikes, innovative dikes (that can be overtopped but do not breach), multi-functional dikes (combining flood defence with other spatial functions such as parks, roads, housing etc.) (Tennekes et al. 2013). Apart from this improving existing drainage channels) or retaining water upstream of the area to be protected (rather than keeping it out) are options as the German and Dutch Room for the River programmes show (Rijke et al. 2013).

Two dominant features seem to characterise the implementation of flood defence. First, due to the public goods character of flood defences, governments at different levels are often key actors, being responsible for their construction, financing and maintenance, Second, discourses on flood defence are highly technical. The design of flood defence structures is based on calculations using estimated probabilities and taking into account factors as population density and potential economic damage (Immink 2005). As a result, flood defence has become de-politicised in many cases. Flood protection standards in The Netherlands are for instance formalised in law, but not discussed with the public. Despite the presence of distributional effects, normative choices (different protection levels, divisions of costs and benefits, risk acceptability) often remain implicit. Examples have been documented where these distributional effects have been made more visible (Warner and Van Buuren 2011; Van Buuren et al. 2013) and politicisation occured. The creation of an emergency retention storage in the Dutch Ooijpolder is a case in point. Inhabitants of this polder successfully resisted this plan, the moment they became aware of it (Roth and Warner 2007).

An overall picture seems to emerge in which the question of who gains and who loses with which types of flood defence measures is becoming more diverse (Penning-Rowsell and Pardoe 2012, 2014). Hence, a key challenge of implementing flood defence can be interpreted as coordination the gap between water system management and society at large by explicating societal debates on flood protection (Mees et al. 2014b; Buchecker et al. in press), employing adequate communication mechanisms (Keller et al. 2006) and providing solutions for dealing with the distributional effects of measures on people, nature, landscape and cultural heritage (Wiering and Driessen 2001).

Mitigation strategies take as a starting point that areas can occasionally be flooded. Examples of this FRMS are the construction of floating houses, urban green infrastructure, and adaptive building.

In general flood risk mitigation is developed and implemented on a local or regional level. While it is widely acknowledged that mitigation (and non-structural measures) are potentially more efficient and sustainable solutions to flood risks than flood defence (Bergström 2006;; Greiving and Angignard 2013; Green 2014), there are no robust requirements for flood risk mitigation measures at the EU or the national level (Van Rijswick and Havekes 2012). Mitigation measures tend to be informal and non-binding. Local governments tend to avoid the implementation of such measures (Neuvel 2009; Heintz et al. 2012). Mainstreaming flood management in future urban planning by adjusting national building codes (Spence 2004) has often been pinpointed as problematic (Fratini et al. 2012; Hartmann and Driessen 2013; Uittenbroek et al. 2014). The WMO (2012) indicates this might be due to the difficulties urban planners have to encompass the vulnerability of areas to flooding.

Measures found in the literature that could stimulate the implementation of flood-risk mitigation strategies include concluding voluntary multi-stakeholder agreements (De Moel et al. 2009), encouraging responsibilities of local authorities (Thaler and Priest 2014) and stressing the local economic benefits of mitigation measures to the community (Holub and Fuchs 2009). In the aftermath of flood events, governments could use arising ‘windows of opportunity’ to make infrastructure and houses (more) flood-proof (Powell and Ringler 2009; Field et al. 2012).

Overall, creating more flood risk awareness and acceptance is key for enhancing the acceptance of new, more resilient urban development (UNECE 2009; Mees et al. 2014a; Fratini et al. 2012; Spence 2004).

Flood preparation includes organisational activities and measures taken in advance to ensure an effective emergency response (Caragliano and Manca 2007).

Key actors in flood preparation measures are – on one hand – the meteorological and hydrological services that prepare forecasts, and on the other the actors that are responsible for developing a flood warning. Another group may be responsible for disseminating the warning to the end-users and also for responding to requests from the end-users for further information. Emergency and social services on their turn have the tasks of taking appropriate actions based on the warning issued. They have to develop and implement evacuation plans.

A communication protocol has to be developed in order to adequately inform the people., since the decision-making process for mass evacuation is influenced by great uncertainties and consequences. A clear formalisation of responsibilities can help to reduce these uncertainties. It must be clear who has to play what role, who is in charge and who is accountable. When it comes to flood preparation, communication is crucial to strengthen awareness and to motivate the population at risk to be prepared for an emergency (Hagemeier-Klose and Wagner 2009).

Risk awareness, key to adequate flood preparation, tends to be low, especially as Kolen and Helsloot (2014) argue in cases in which the return period of a flood is far greater than a lifetime. Such a low public awareness restricts decision-makers when prioritising emergency planning (Ten Brinke et al. 2008). Trust in the government on the other hand causes people to refrain from making private preparations.

Aim of this FRMS is to facilitate a good and swift recovery after a flood event has taken place. Measures include the development of compensation and insurance systems as well as the preparation of reconstruction or rebuilding plans.

Flood recovery can be a rational FRMS, as it might be too expensive to implement structural flood defence measures. This will be the case for less populated or economically more marginal areas. As there will always be a chance of dike breaches, piping or overtopping, recovery also has to be organised for protected areas (Klijn et al. 2009).

Rebuilding costs can be very high. Key question in cases like these is of course who must pay the bills (Penning-Rowsell and Pardoe 2012).

Profit-driven private insurance systems will not be solvent to do so in cases of serious flooding (Botzen and Van den Bergh 2008) and do exclude those unable to pay their tariffs. Either governments or civilians have to deal with the non-insured risks (Aerts and Botzen 2011, p. 1046).

In a comprehensive disaster insurance program different actors can be discerned: property owners (self-insurance), insurance companies (private insurance), reinsurers and capital markets (risk transfer) and state reinsurance (state funds, multi-state pools) (Kunreuther and Pauly (2006, p. 112). In Germany (Thieken et al. 2007), the UK (Penning-Rowsell et al. 2014) and France the insurance industry plays a role in flood risk management (Botzen and Van den Bergh 2008). In the Netherlands however, only one private flood insurance scheme (in a pilot phase) exists. In cases of flooding a national disaster fund would pay the victims (Aerts and Botzen 2011).

The overall challenge for actors that want to put more emphasis on flood recovery is to clarify which risks will be covered by governmental funds and which by privatised flood insurances, and whether buying private insurance is voluntary or mandatory (Kunreuther and Pauly 2006). The development of private flood insurance systems may be easier if all natural disasters insurances can be combined into a single policy, as it is more likely that a property owner will consider purchasing insurance because the likelihood of some loss is above a threshold level of concern. If all natural hazards were included in a homeowners’ policy the risk might be large enough to get the consumer’s attention. Insurance companies can offer landowners reduced premiums if they take measures themselves to reduce their risks (Hartmann 2009, 541). The latter also implies that working forms of cooperation (e.g. partnerships) have to be developed between residents, (re)insurance companies and governments (Kunreuther and Pauly 2006).

Insurance companies somehow have to address moral hazard problems (like moving unwanted furniture to the basement in order to have the insurance companies pay for their replacement). Furthermore, some residents will not buy policies even at subsidised rates because they believe disasters would not happen to them. This again points to the necessity of risk communication in clear language (Kunreuther and Pauly 2006).