Flood risk management and ‘fairness’: aspirations and reality

Flood risk management in United Kingdom has been going through a process of rapid change in the last decade or so, no doubt spurred on by a series of very serious floods since the year 2000. These changes affect flood defence and non-structural flood risk management measures alike, and involve a degree of devolution from central government to local communities and regional organisations, as central government seeks to shed responsibilities for policy implementation. This paper discusses three case studies concerning flood defence, property level protection, approaches to social justice. The results show a different pattern in each area, with flood defence moving somewhat towards a Rawlsian approach, but flood insurance and property level protection showing signs of both inefficiency and poor penetration, respectively, particularly with regard to low income residents, especially those in social housing.


Introduction
Those concerned to manage and hence reduce flood risk commonly have aspirations to be fair, and treat all those at risk and contributing to risk reduction equally in an open and transparent way that maximises social wellbeing and solidarity, in the face of both risks now and risks into the future.Such aims are often articulated in policy documents, Government statements and through the work of Non-Governmental Organisations.They also involve consideration of relevant populations both during flood events, in the recovery period afterwards, and through the investment of state and private resources targeted at risk reduction measures, both engineering and non-engineering.
In this paper these issues are discussed with regard to three examples.First, we look at investment in traditional flood risk management (FRM) measures in the UK, and the impact they have on deprived and non-deprived communities.This appears to show a marked shift in attention towards the former, and away from the latter, as government investment is targeted at those areas of financial deprivation which are also at flood risk.This is not an absolutely clear picture, because at the same time the government is prioritizing investment in a major scheme to the west of London, the Thames Scheme, where substantial investment has been given without consideration, it would appear, to the social composition of the area and its relative affluence.
Secondly, we looked at flood risk management for those in social housing and the provision of propertylevel-protection measures.Such measures are unaffordable to those on low incomes in social housing, yet this is one policy option that is being promulgated widely in the absence of sufficient government funds for major flood risk reduction measures of the traditional kind.Leaving those who are already vulnerable to seek protection from their own resources appears to be a perverse policy, with few chances of significant success.
Thirdly, we look at recovery and insurance that promotes recovery, and a new proposal in the UK to subsidise flood insurance (through the reinsurance arrangement termed Flood Re), even for those occupying very large houses which are worth considerable sums of money.
The whole question of subsidising flood insurance is an important one in justice terms, because it means a cross subsidy from those not at risk to those at risk (or from those at low risk to those at high risk), and this has become more transparent recently owing to government policies to make flood insurance affordable and available to anyone, even those at high risk.@ However, two broad characteristics of social housing units limit the ability of their occupants to engage with the flood risk management policies, in particular the widely advocated policy of encouraging property-levelprotection measures (PLP).This policy seeks to transfer the responsibility of flood protection to the residents themselves, for example through their implementation of coping mechanisms (e.g.moving belongings), immediate adaptations (e.g. using sandbags), or future adaptations (e.g.raising the height of the front door threshold) to cope with flood risk [14].

6RFLDO MXVWLFH
7KH Whilst a policy stipulating increased tenant responsibility for their flood risk is in theory an effective manner to increased flood preparedness for the UK at a low cost to the government, it is only effective if the tenants themselves engage with this policy.To ask social housing residents who are often the most deprived individuals in the country to engage with this policy, as though they are equally prepared and able to as those who are much more capable and affluent, shows a severe lack of understanding towards the concept of differentiated vulnerability.Given their characteristic limited income and educational levels, and the confused status of autonomy of residents in social housing, it is likely this policy will have limited chance of success.
Transformational adaptations, which are more significant actions like moving housing to avoid the flood risk [18], are unlikely to be supported by the government or by the tenants themselves given the strict allocation criteria for social housing which means that moving to DQRWKHU DUHD FDQ ORVH WKH WHQDQWV ¶ OLNHOLKRRG RI EHLQJ offered affordable state or housing association promoted housing.

Flood insurance post 2016 in the UK
Flood insurance is near universal in the UK, provided by private companies [19].But from April 2016 the UK DGRSWHG D QHZ µSRROLQJ ¶ DSSURDFK )ORRG 5H implemented into UK law by the Water Act 2014 [20] and additional Regulations [21], has the aim of promoting the continued availability and affordability of insurance premiums in flood risk areas.Flood Re applies a (graduated) premium cap to household policies thereby limiting the amount paid by an individual for flood insurance.Insurers then cede these policies into the Flood Re scheme and any flood losses that occur are paid from the industry-wide pool.The £180m annual funding for the pool is generated from the premium income of those properties entered and also from a levy which is included on all domestic flood insurance policies, which are now formally subsidising the insurance of properties in high flood risk areas.Flood Re is only meant to be in place for the next 25 years, supposedly to give both the government and households time to reduce flood risk.Over this period, it is proposed that the capped premiums will increase, thereby phasing out the cross-subsidy and returning flood insurance to a purely market based situation [22].
Flood Re raises a number of social justice issues.The first relates to which properties are eligible.Only private households are included [23]; businesses are not included, nor landlords with multiple properties, so that most leasehold properties are excluded New properties built after 1st January 2009 are also excluded, to recognise that a critical component of UK FRM is to prevent the increase of flood risk.But problems relating to flood risk awareness and disclosure may mean the burden is passed to unsuspecting homeowners, rather than remaining with developers, who retain no liability.
Maintaining the affordability (and availability) of flood insurance appears to be a positive step, but there is the question of who benefits.Flood Re is specifically designed to assist those householders with high premiums for flood cover and as such it is only these households in higher risk areas that benefit: those at higher flood risk are being subsidised by those at low or no flood risk.$OWKRXJK WKLV VLWXDWLRQ GLG H[LVW SULRU WR WKH VFKHPH ¶V implementation, Flood Re has formalised this relationship and Penning-Rowsell and Pardoe [24] raise questions about the fairness of the approach.
Within the approach there is little consideration of the ability to pay for flood insurance as a criterion for receiving premium assistance.The way in which Flood Re caps premiums and distributes the risk amongst all policyholders means that low income households who are not at risk are cross-subsidising high income, at-risk households who could very probably afford the riskreflective price.The capped premiums do show some graduation according Council Tax property value ³%ands´ LH YDOXH UDQJHV and those residents in lower value properties pay less for their insurance than those in more expensive properties.Low value ³Band A´ properties would pay £210 towards the flood component of a combined flood policy, whereas a mid-YDOXH ³Band E´ property has a premium cap of £330.This is acting as very crude proxy for whether a property owner may be able to afford to pay.However, this graduation is in part related to the additional cover required by more expensive larger properties: any claims will be higher.Original plans for Flood Re excluded the most expensive 1% of properties (³Band H´), recognising that these occupants should not subsidised [25].However, the scheme as implemented did include these properties, although the capped premium is more than double the next lowest band at £1200 [26,27], but it would be considerably higher without the subsidy.Furthermore, Flood Re does nothing to address the wider issues of affordability of insurance and there has not been the inclusion of any mechanism to enable those who are not able to afford any kind of insurance to access flood cover.
A counter to these social justice concerns is that Flood Re is only intended to be a transitional arrangement and that over time this cross-subsidy will be removed, at least for the higher value properties.However, the first transitional plan [22] provides few details about how the removal of the cross-subsidy will work in practice.For instance, unless flood premiums are able to be lowered due to risk reduction measures it is likely that affordability will remain a key issue and may lead to a reduction in insurance penetration.Additionally, as indicated above, those on lower incomes are less likely to be able to afford to take proactive PLP measures to adopt risk reduction measures, as well as being less likely to be able to afford higher premiums with the transition to riskrelated premiums [28,29].Therefore, unless there is government intervention either better to enable disadvantaged households to adopt flood mitigation, or the cross-subsidy continues for these low income households after the proposed end of the scheme, it is likely that greater numbers of properties will become uninsured in the future, occupied by those less able to afford to recover from flooding.

Assessment and conclusions
The last decade has seen very significant changes in flood risk management policy in the UK [30].Many of these changes have been directly or indirectly a result of severe flooding in 2000, 2007, 2013/14 and 2015/16.Policy changes have affected both the non-structural and the structural dimensions of flood risk management measures, as the government and other stakeholders have seen the necessity to react to a change in flood risk regime and to a financial state of affairs whereby governments are seeking to restrain their expenditure in this area, given the reluctance to raise taxes and extreme budgetary pressures elsewhere, for example from the health sector.
The essence of the changes flood risk management policy has been to devolve responsibility to local and regional actors, including individual households, rather than central government being the prime mover for risk reduction.
There are some contradictory elements with regard to social justice and fairness in this respect, as characterised by our three examples discussed above.Flood defence expenditure appears to be moving towards quite an implicit Rawlsian approach, at least in part, giving a distinct advantage to protecting householders in financially deprived areas and making the implementation of large flood risk management measures in more affluent areas difficult to fund.
On the other hand property level protection measures undertaken by individuals in social housing -also characteristic of socially and economically deprived locations ± appears to be completely neglected, thus adding to the vulnerability of the occupants and restricting their ability to recover from flood events.
At the same time, the new Flood Re proposals for flood insurance continue to subsidise for the next 25 years those householders living in floodplain areas, including subsidising those almost certainly well able to afford market prices for the premiums they pay for the cover provided.The scheme also appears to continue to create a distinct disincentive for individual householders who have their insurance subsidised in this way to reduce the risk they face by taking some actions themselves.
Both the social housing and the flood insurance situations are distinctly non-Rawlsian in character, but nor are they egalitarian, nor necessarily utilitarian.Indeed, the flood insurance proposals cannot be seen as utilitarian since they are not characterised by the search for efficiency that that philosophy espouses: the latest assessment of Flood-Re suggests that its net present value is negative [31, 32]!The most charitable view of the situation with regard to social housing is that it has not been considered a significant flood risk management issue in the past, rather than the situation being deliberately designed to penalize those living in such locations.Whilst we recognise that flooding itself is not ³IDLU´ VLQFH LW DIIHFWV RQO\ D VPDOO SURSRUWLRQ RI WKH population [2], movements in the policy arena that exacerbate rather than minimise unfairness would not seem to be a sustainable or sensible way forward.

Figure 1 .
Figure 1.Current and future spending on flood and coastal defence against the latest assessment of need as SXEOLVKHG LQ WKH (QYLURQPHQW $JHQF\ ¶V ORQJ-term investment strategy [32].