The allusive market: insurance of flood risk in neoliberal Britain

Abstract Launched in 2016, Flood Re is a government-supported scheme for flood-risk insurance in Britain that is intended to pave the way towards an eventual ‘free’ market featuring risk-reflective pricing. This paper introduces the concept of ‘the allusive market’ to denote the figurative work that the market vision performs in this context. Alluding to the merits of what is in reality a highly implausible market-based future for flood insurance releases the government from having to substantively address intractable problems associated with the financial risk of flooding in the present: the market will come to the rescue. A risk-management crutch, the allusive market engenders contemporary policy paralysis, occasioning in turn the worsening of the very problems that the market is being relied upon, eventually, to resolve.

launched a new vehicle designed to help secure the availability and affordability of home insurance for the approximately 350,000 British households located in areas at high risk of flooding: Flood Re. This new arrangement enables insurers to offer most such householdsthere are exclusionsaffordable buildings and contents cover, which is to say cover priced at a level that does not reflect the actual flood risk to which they are exposed. Flood Re provides this subsidy mechanism by offering insurers reinsurance (hence, Flood Re): in the event of a claim, Flood Re will reimburse the relevant insurer specifically for the flood element of that claim. Industry-owned and run on a not-for-profit basis, but directly accountable to Parliament under the terms of the 2014 and 2015 legislation that mandated its establishment, Flood Re is funded by an annual levy raised from all insurers writing home insurance in the United Kingdom. By September 2017, some 60 insurers were offering policies backed by Flood Re, and 142,000 such policies had been written (Lucas, 2018).
Flood Re is not a market solution to the longstanding problem of providing affordable insurance to high flood-risk households. But, crucially, it is not entirely a non-market solution, either. It is, precisely, a waystation to a market solution. The government, the insurance industry and the enabling legislation all posit Flood Re as a temporary, interim scheme that will eventually pave the way for a 'pure' market in home insurance, where the price of insurance does reflect underlying flood risk, to take shape. In fact, it is on exactly this transitional basis that Flood Re has been rationalized: what it isan arrangement structured by subsidy, and universally acknowledged to be imperfectis justified by what it will become. As the company (Flood Re, 2016a, p. 11) itself explains: Flood Re has been designed to provide temporary support to the insurance market for properties at high risk of flooding. At the end of 25 years from the date the Water Act 2014 received Royal Assent (May 2014), Flood Re will have been wound up and the subsidy provided through the scheme removed.
And left in Flood Re's wake? 'A market with risk-reflective pricing should then be in place'.
Focusing on this idea of institutional 'becoming' or metamorphosis, the present paper describes and analyses Flood Re as an example of a particular modality of market-led governance and risk management, relying on a particular market figuration or trope: namely, the market as the answer, eventually. This trope, and its mobilization and effects, constitutes the paper's main theoretical and empirical concern. I refer to it as the allusive market. Ever since Flood Re was initially mooted, discussion of its role and structure has consistently been framed by references to what will succeed it; it has never been thought of, at least publicly, as a permanent solution. Yet, the nature of the market that will supplant it has never been specified. That a 'genuine' market will follow is simply taken for granted. Nowhere is this combination of presumption, on the one hand, and the lack of detail, on the other, clearer than on Flood Re's own 'explainer' webpage for insurance industry partners. 'In 2039', it announces, as if remarking that night follows day, 'the Flood Re scheme will end and there will be a free market for flood risk insurance'. 1 There just will. But this presumed market lacks form. As invoked by Flood Re's designers and main institutional stakeholders, the market that will come afterwards is merely symbolic, figurative. Alluded to rather than actively delineated, this market is, in sum, allusive.
The paper explores various dimensions of this allusive market, through the course of four main sections. The first section situates the paper's principal conceptual arguments in relation to existing heterodox literatures on markets, their idealization in neoliberal thought, and what has been termed their 'performativity'. The second section provides a different type of context by considering the range of different international approaches to insuring household flood risk. A capsule history of flood risk insurance in Britain up until the creation of Flood Re is provided. Flood Re is the focus of the third section of the paper, which, in examining it as a means to an endthe precursor to the ultimate objective of a self-regulating (albeit allusive) markettheorizes it loosely in the Gramscian terms contained in the paper's epigraph. In Britain, the olda non-market approach to flood risk insuranceis dead, or dying. The market is the new, preferred approach. But, for reasons that I elucidate, the new cannot yet be born. The fourth and final section of the paper focuses on Gramsci's interregnum, the here-and-now between the death of the old and the deferred arrival of the new. My main interest is in the likely effects in Britain of having implemented a transitional flood insurance solution, and especially of framing that solution as a mere segue to something different and better. What are the likely consequences of saying that there will be a market but not actually putting it in place? Of alluding to, but not enacting, an eventual market solution? The answer, I suggest, is the same as Gramsci says it always is when the old is dying and the new cannot be born: crisis.

The allusive market
Long an ontological and conceptual centrepiece of mainstream economics, markets have in recent decades become an increasingly important object of enquiry for various heterodox economic traditions including economic sociology, economic anthropology and economic geography. The economic sociologist Patrik Aspers (2011, p. 4) defines a market as 'a social structure for the exchange of rights in which offers are evaluated and priced, and compete with one another'. This relatively straightforward definition suggests a relatively straightforward phenomenon. But the reality, of course, is different.
Indeed, prominent scholars have made compelling claims that the market, for all the attention dedicated to it, remains in crucial respects elusive. It is, on the one hand, theoretically elusive, Greta Krippner (2001, p. 778) arguing that notwithstanding sociologists' and others' efforts to 'delineate features of markets', the market per se 'has tended to elude researchers as a sociological object'. The market, Daniel Miller (2002) asserts, and perhaps more contentiously, is also empirically elusive. What he means by this is that real-world mechanisms of exchange rarely if ever look anything like textbook market mechanisms. Conceptually underspecified, in other words, the market is also empirically indeterminate. Scholars looking for it on the ground have struggled to find substantive evidence of it. If 'the market' exists, in short, it is, says Miller, only as an ideological model.
The argument of this paper is that the market is not only elusive but also, in certain circumstances, allusive. It is something that is deliberately alluded to; and alluding to itinvoking it, projecting it, and, in particular, promising it, but without ever actually implementing itdoes important political work. The paper makes this argument through a consideration of the market for flood risk insurance in Britain. This 'market' is not currently a market, nor considered as such by insurers, regulators or the government. It is scaffolded by subsidies, and is thus not 'free' in the sense of featuring prices thatas they would in a 'real' insurance marketfully or even approximately reflect risk. But a proper, functioning market for flood insurance is, the UK government says, possible. It is, moreover, the future: such a market will, eventually, materialize, at least according to the governmenton current projections, in 2039. 'The market' thus figures allusively in government discourse on flood risk and flood risk management, as somethingfigurative rather than literal, promised rather than produced, and always lacking detailed specificationthat will, one day, come into being. This allusive market is, perhaps above all else, possible. This possibilistic quality matters, I argue, because presumed possibility is inherently political. Insisting that a market for flood insurance is possible, and alluding to its future crystallization, serves to obviate the need for pursuing in the here-and-now sustainable, non-transitory, non-market-based approaches to flood risk management. The market, in sum, will save the day.
The trope of the allusive market, it should be noted, conforms seamlessly with the neoliberal playbook. Neoliberalism is notorious for 'failing forwards' (Peck, 2010, p. 6). Not only does it have a resolutely 'constructivist' orientation, according to which 'Any problem, economic or otherwise, has a market solution, given sufficient ingenuity' (Mirowski, 2009, p. 439). But for neoliberals, the market is also the answer, paradoxically, to markets' own failures, from which it can instinctively learn. Thus the 'manifest inadequacies' of neoliberal market experiments merely animate 'further rounds of neoliberal invention' rather than a step backwards (Peck, 2010, p. 6). Indeed, the quasi-evangelical conviction that 'the market (suitably reengineered and promoted) can always provide solutions to problems seemingly caused by the market in the first place' is nothing less than a neoliberal sine qua non (Mirowski, 2009, p. 439). In this regard, neoliberalism is a philosophy of futurity, forever conveying an indication of something to come: a better, truer, 'real' or 'free' market society, unimpeded by government intervention. If neoliberalism is indeed fundamentally promissory, then the allusive marketthe seductive promise at its core, and which sustains its mythologyis arguably its cardinal trope.
There are also notable parallels between the issues examined in this paper and those examined within the literature on market 'performativity'. As I have said, I am interested here in the relationship between socioeconomic realities on the one hand and, on the other, ideas or claims about those realities, or more specifically about the form they will eventually take. This, generically, is the central concern of the performativity literature, too. It is now two decades since Michel Callon (1998) influentially posited economic theory as existing in a performative relationship with actually-existing markets: helping them dance to its own tune, and perhaps even to come into being, rather than simply passively describing them.
Nevertheless, there are significant differences between the work performed by the trope of the allusive market and the performative work examined by Callon and his followers. These differences relate not so much to the distance between the allusive marketthe idealized market, that is to say, that is alluded toand the actually-existing (non-or quasi-market) reality of present-day flood risk insurance in Britain. After all, the literature on performativity is itself now replete with examples of economic theory not successfully shaping the world in its own image: perhaps most notably, Donald MacKenzie's (2004) discussion of economic models misfiring and, in the process, producing, 'counterperformatively', inverse effects. Rather, the key difference between the allusive market and the models of the performativity literature is that the former is merely allusivesomething promised rather than performed, whether with positive or inverse effects.
As we will see, the principal effect of the allusive market is not to perform but rather to precludespecifically, by virtue of the insistence that a market will ultimately take shape, to preclude meaningful government consideration and delivery of durable non-market flood risk-insurance solutions. If alluding to and promising an eventual market outcome for flood risk insurance has any performative effects, it is to seed not a market per se but what Judith Butler (2010) felicitously calls 'the market presumption'. 'A series of discursive and non-discursive practices and institutions', Butler writes of this generic presumption (p. 148), 're-constitute the idea of the market as [a potential] reality'. Such a presumption that there can, and at some point in the future will, be a market, produces, in the UK flood-risk management context, effective policy paralysis. And it is precisely in the context of this policy paralysis that a great variety of morbid symptoms is liable to appear.
Finally, it is worth registering family resemblances between the allusive market of the present paper and the market allusions explored by Luc Boltanski and Laurent Thévenot in their influential On justification (2006). Boltanski and Thévenot identify the discourse of 'market worlds' as one of six main logics invoked in contemporary capitalist society to justify and rationalize various forms of social action and worth. Here, people and institutions seek to legitimate themselves and their actions by reference to market logic: this or that action makes sense, and this or that person has value, because they and their actions conform to and reinforce a social market order. There is certainly something of this justificatory reasoning in the market allusions examined in what follows, which also inherently privilege market forms of organization. But where Boltanski and Thévenot discuss a generic market logic, the British flood insurance case concerns an actual, if hypothetical, market: a market that will, in theory, materialize, and thus which eliminates the need to find a non-market solution to flood risk insurance, as opposed to a nebulous market philosophy that serves as a yardstick of legitimacy.

Flood risk and flood insurance
Alongside earthquakes, hurricanes, volcanic eruptions and tsunamis, floods number among the most recognizable types of natural disaster. In Britain, they are easily the most common form. Societies can and do respond to the threat of flooding in various different ways. Through the actions of individuals, communities, and both private and public institutions (including governments), they can attempt to minimize the frequency and intensity of flooding by reconfiguring the geophysical landscape; they can attempt to minimize the damages caused by floods by planning, building and living with flood risk in mind; and they can attempt to provide relief to those that suffer from any such damages. This last component of the social response to flood risk is where flood insurancecoverage against property loss from floodingenters the picture. But the key point is that flood insurance (or, more generally, mechanisms for managing the financial risk of flooding) is always just one element of a wider bundle of social responses. Furthermore, these various responses not only typically involve an array of different actors at different scales, but, as we will see, they are connected to one another, not least insofar as changes to one type of response mechanism tend to have implications for responses of other kindsif a community builds impenetrable levees, for instance, this will clearly affect where people feel comfortable living and how important they feel it is to protect against potential property damage. 'How society responds to [flood] risks', then, is, as Surminski and Eldridge (2017, p. 415) observe, 'not simply a question of engineering; it is a rather complex area, with political, economic, social and environmental dimensions'.
We see this in Britain, where a veritable panoply of different bodies have formal responsibilities for managing flood risk. Overall policy responsibility lies with the Department for Environment, Food & Rural Affairs (Defra), which provides funding for flood risk management to a series of other bodies charged with delivering its overall policy: principally, the national environmental agencies of England, Scotland and Wales, alongside local authorities, highway authorities, internal drainage boards and several others. In 2013, Defra summarized its (and the government's) headline policy as follows: 'tackling flood risk whilst avoiding inappropriate new development in the flood plain' (Defra, 2013a, p. 2). Or, as the Foresight Land Use Futures Project report had put it, more fully, three years earlier: In broad terms, government policy for flood risk management adopts a riskbased approach [that] focuses expenditure where it will best achieve a set of outcome measures associated with economic benefits, number of households protected, protection of households in deprived areas, protection of nationally important wildlife sites and achievement of biodiversity targets. … Guidance from Government on development and flood risk states that inappropriate development should be avoided in areas at risk of flooding, and development should be directed away from areas at highest risk. (Government Office for Science, 2010, p. 155) Different countries illustrate different approaches to the matter specifically of managing the financial risk of flooding (Defra, 2013b, pp. 54-55;O'Neill & O'Neill, 2012, p. 4). At one end of the spectrum are those where the state, and thus the taxpayer, essentially bears all the risk. The Netherlands, where more than two-thirds of homes are located in at-risk areas, is one such: flooding is generally excluded from home insurance policies, and the state provides compensation for losses. At the other end of the spectrum are market-based approaches, where individual households bear their 'own' risk and insurers set private policy premiums on the basis of the estimated probability of loss. Australia and Ireland are the only significant existing examples of this approach. Other countries, meanwhile, operate a range of hybrid approaches involving both the state and the insurance industry in some sort of combination. In France, for example, there is a partnership between the two: The insurance industry collects a compulsory premium for natural disasters that is standard in policies and which is charged regardless of the level of risk. The state acts as the re-insurer and hence guarantees payments as the insurer of last resort. Insurance payments are made in the event of a state announcement of a state of natural catastrophe. (O'Neill & O'Neill, 2012, p. 4) In the United States, many households buy policies inclusive of flood cover in the private home insurance market, but such cover is generally excluded in high-risk areas, where households can instead make use of the federal National Flood Insurance Program (NFIP) if they live in communities that adopt plans to reduce future flood risks.
As Stephen Collier (2014) and others have observed, the approach that a country takes to managing the financial risk of flooding is typically reflective of prevailing political-economic orthodoxies. For insurance is, in Collier's words, a 'political technology'. State-based approaches such as the Netherlands' are rooted in principles of solidarity. They spread risk across households of varied circumstances, who therefore cross-subsidize one another. Marketbased approaches disclose an altogether different 'moral economy of disaster a new calculus of loss, compensation, and responsibility'. Here, the individual homeowner must 'weigh the benefits of owning property in a flood prone area against the costs of insuring that property' (Collier, 2014, p. 275). Individual 'responsibility' is paramount; it is every woman or man for themselves. This risk privatization model is fundamentally (neo)liberal, and it is no coincidence, of course, that Australia and Ireland are neoliberal poster children. All approaches to flood risk insurance, meanwhile, have their critics, and these criticisms are important to bear in mind when we turn, presently, to the British case. Where the government is involved as insurer of last resort (in either hybrid or state-only systems), the charge is of moral hazard: developers have little incentive to stop building on flood plains, and homeowners have little incentive to flood-proof their properties, because they know the government stands ready to meet the cost of damages. In market-based systems, by contrast, the charge is of financial exclusion: flood insurance for homes at high risk is unaffordable or simply unavailable, with low-income households being disproportionately affected.
Historically, the British approach to managing financial flood risk has been rather unique. From the early 1960s until 2016 and the launch of Flood Re, there was, essentially, a 'pact' between the government and the insurance industry. The former would build flood defence infrastructure and prevent development in high-risk areas; the latter would provide cover for flood risk (which is a requirement of UK mortgage lenders), even for high-risk households, and, crucially, at least in the pact's original form, at affordable rates. But for a number of reasons this pact has in recent decades come under increasing strain, and it eventually disintegrated. It is in the context of, indeed because of, this disintegration that Flood Re was conceived and designed.
Flood cover was available from some British insurers from as early as the 1920s. But coverage was always extremely patchy; most households were not insured. 'Then in 1961, to stave off Tony Benn's threat to nationalise insurance when the Labour Party next came to power', writes David Crichton (2005, pp. 7-8), 'the British insurance industry voluntarily agreed amongst themselves to guarantee to provide cheap flood insurance for all households and small shops, regardless of where they were situated'. In other words, this was explicitly not a market: cross-subsidy, flowing from low-to high-risk households, was baked into the system from the start. In return, the government pledged to make the aforementioned investment in flood defences while limiting floodplain development. The pact between the two sides was known as the Gentlemen's Agreement.
The Agreement held until the early 2000s. The government would likely have been happy for it to remain in place in perpetuity. But the insurance industry had become increasingly disaffected. The main reason was that it did not see the government keeping its side of the bargain. As Crichton (2005, pp. 8-9) explains, 'more and more new houses [were] built in flood hazard areas, often against the advice of the Environment Agency'and insurers nonetheless had to provide affordable flood cover. This was moral hazard writ large, only the guarantee of payout for damages was providedincreasingly unwillingly by industry rather than the state. Again, Crichton: If the market had been free to apply prudent underwriting standards, many of these houses would not have been sold because purchasers would not have been able to afford the insurance premium and therefore would not have obtained mortgages. In any event, after forty years of the market distortion caused by the guarantee, insurers [found] themselves faced with the situation of a large and growing number of houses at risk from flood, as planners and developers [had] taken the continuing availability of cheap flood insurance for granted. (2005, p. 8) The straw that broke the camel's back was large pay-outs for floods in the late 1990s (O'Neill & O'Neill, 2012, p. 4). The industry, preoccupied with profit margins, had had enough.
So, although the historic pact between industry and government was renewed in 2002, with the Gentlemen's Agreement being replaced by the so-called Statement of Principles, the latter was only ever intended to be a temporary solution (much like Flood Re, one might note). And the government, which committed to taking action to reduce flood risk by investing further in defences, had conceded notable ground. For one thing, some homes, albeit a very small number, were rendered uninsurable, since the industry pledged to cover only homes with a probability of being flooded in any single year of 1-in-75 or less, or where flood defences planned for the next five years would bring the risk down to that level. For another, the premiums and excesses charged by insurers would now be allowed to reflect different levels of flood risk; affordability of cover was no longer guaranteed. When a revised iteration of the Statement of Principles was agreed in 2008, additional exclusions were incorporated. To discourage new development in high-risk areas, properties built after 1 January 2009 would not be covered by the industry's commitment.
Did the Statement of Principles herald the arrival in Britain of a market-based approach to flood risk, à la Australia and Ireland? Insofar as risk-based pricing of flood policies was now permitted, it certainly appeared to presage marketization. Yet the industry was still under an obligation to make cover available to most properties, including some at very high risk. Faced with a choice between, on the one hand, removing subsidies, and thus charging essentially unaffordable prices to high-risk households, and, on the other, retaining subsidies that very few low-risk households even knew they were paying in the first place, the bulk of the industry opted for the latter. This decision was reinforced by the fact that in the early 2000s, when the Statement was agreed, insurers' flood-mapping technologies remained relatively crude, militating against riskreflective pricing (Crichton, 2005, p. 8). As such, subsidies, amounting to an estimated £150 million per annum (Defra, 2013b, pp. 12-13), continued to course through the system. As recently as 2011, 78 per cent of insured homes in areas of significant flood risk were still being subsidised by homes in areas of lower risk (O'Neill & O'Neill, 2012, p. 4). The average 'under-pricing' for these subsidised homes was estimated to be approximately £430, with one in seven being under-priced by over £1,500 (Flood Re, 2016a, p. 21).
But while Britain thus still did not have a flood insurance market per se, the old, non-market approach was well and truly dying: 22 per cent of high-risk insured homes, after all, were now paying premiums reflecting actual flood risk. Insurers, in short, were becoming increasingly bullish in taking advantage of the pricing freedoms that the Statement of Principles afforded them. In this they were aided by the availability of more sophisticated flood mapping models (Defra, 2013b, p. 13;c.f. Collinson, 2017). And as the prices 'offered' to highrisk households rose, the number facing exclusion rose in lockstep. The ABI estimated in 2012 that as many as 200,000 households, especially those previously flooded, now faced difficulties obtaining insurance because the cost was too high or the risk simply deemed uninsurable (Bachelor, 2012). The widespread flooding of the summer of 2007 had made it clear that the problem was not just a residential one, either: in Hull, where two months' worth of rain fell in two hours, the local council was left with a colossal repair bill having decided not to insure schools, council houses and other public buildings for water damage because the insurance excess was considered prohibitive (Gray, 2007;Woods, 2007). It transpired that Hull was not alone in this decision: over half of UK councils were self-funding against storm and flood damage (Gray, 2007).
All of this was compounded moreover by the fact that building on flood plains had continued unabated in the face of growing demand for new housing, with local authorities complaining that 'the Government are putting so much pressure on them to build a given number of houses that they have no choice but to build them in green-belt and flood-risk areas' (HC Deb 26 February 2014, c339). 171,700 new homes were constructed in flood-risk areas between 1997 and 2010, peaking at 14,500 in 2006 (Martin, 2014); being cheap, such flood plain land was (and is) often used for social housing, 'where the residents may not be able to afford to buy insurance [and] did not have the resources or influence to agitate for improved protection' (Crichton, 2005, p. 14). When James Meek visited Tewkesbury in Gloucestershire in the wake of the 2007 floods, locals ridiculed the government's guidance for local authorities regarding limiting building on flood plains as 'the biggest confidence trick by a British government since the Second World War' (Meek, 2008). Development on flood plains not only puts property and people at risk; it also alters the transmission of flood plain flows (Government Office for Science, 2010, p. 161). And, all the while, in the background, heightening the stakes involved in managing the financial risk of flooding, there was the spectre of climate change. In 2009, the Environment Agency (EA) estimated that by 2035, in the absence of a substantive increase in expenditure on flood defences, the number of existing properties exposed to significant risk of flooding in England alone could rise from around 500,000 to over 800,000 (Government Office for Science, 2010, p. 161). The government considers floods to be the greatest threat that climate change poses to Britain.
If, then, in the late 1990s the dissatisfied partner to the pact that had delivered flood risk insurance to Britain since the 1960s had been the insurance industry, in the late 2000s and early 2010s it was very much the government. Where the insurance industry had been concerned about profitability, the government, now dealing with a partner increasingly pursuing profitable pricing under the accommodating Statement of Principles, was concerned about exclusion. As it surveyed the landscape of flood risk insurance in 2013, the government saw Britain heading inexorably towards a market outcome; but it also saw a Britain that was clearly not readyyetfor such an outcome. 'If there was an immediate move to risk-reflective pricing', Defra (2013b, p. 14) warned, 'large numbers of households could see prices rise suddenly and tens of thousands of households could as a result become uninsured'. This was something the government was not prepared to countenance.

Flood Re
The first crucial thing to note about Flood Re, therefore, is that it was conceived and introduced explicitly in order to avoid a free market in flood insurance. Under the Statement of Principles, Britain had been edging in this direction. In the early months of 2013, 'edging' was threatening to become toppling headlong over the perceived precipice, for the revised Statement of Principles agreed in 2008 had a fixed five-year term and no renewal had been agreed between the government and industry. Defra's warning of 'an immediate move to riskreflective pricing', in other words, was far from idle. The urgency of the situation was viscerally apparent when parliament gathered to debate the matter in late March. Dominic Raab, the Tory MP, informed the house that the government and the ABI remained 'deadlocked in negotiations' over what would happen after the Statement expired. And he made it as clear as it could possibly be that the priority was to prevent a free market coming into existence, as it were, by default. 'It is clear that the state needs to be involved in this', Raab said of the immediate future of flood risk insurance in Britain: 'This cannot be left solely to a free market. Even a free-market MP like me would accept that'. Speaking from the other side of the house, the Labour Co-operative MP Gavin Shuker reiterated the point, making a free market sound like Armageddon: Ninety-six days are all that stand between today's near-universal coverage for flood damage and an unfettered free market that will leave tens of thousands of people with homes that are uninsurable, unmortgageable and unsellable -96 days, and the clock is ticking. (HC Deb 26 March 2013, cc1520-21, 1540 And yet the central curiosity of Flood Re is that, simultaneously, it was conceived and introduced explicitly in order to seed a free market in flood insurance. This is the secondseemingly contradictory, certainly paradoxicalcrucial thing to note about it. Flood Re was and is shot through with this particular tension: it is a putative waystation to a free market, whose primary purpose at birth was to forestall the very thing it is ultimately expected to actualize. For there can be no debating the fact that Flood Re embodies a market ideology, notwithstanding its initial market-avoidance remit. To be sure, it represents yet another 'pact' of sorts between the government and the insurance industry, following in this respect in the footsteps of the Gentlemen's Agreement and Statement of Principles. But what distinguishes Flood Re from its predecessor 'partnerships' is that, for the first time, a 'free' market in flood insurance is the explicit objectiveeventually. It is helpful therefore to differentiate between the short and long-term objectives for Flood Re. In parliamentary debate in early 2014, the Liberal Democrat MP Dan Rogerson said the former was to 'deal with the availability and affordability of flood insurance for households at high risk', while the latter was to 'ensure a smooth transition to a free market' (HC Deb 6 January 2014, c135; c.f. Defra, 2013b, p. 17). Thus, as the legal service Practical Law explained, high-risk homes insulated from risk-based pricing by Flood Re would, upon its termination, 'be exposed to the normal rigours of the risk-based market' (Practical Law, 2017, p. 4). But, critically, the intention is that these rigours will not at that juncture translate into uninsurability or unaffordability, which is the outcome that the government feared would result if market forces were unleashed directly upon expiry of the Statement of Principles; rather, as Flood Re itself says, the long-term goal is affordable insurance for high flood risk households in the open market … Its long-term goal is to ensure that the benefits of Flood Re, in terms of affordability and availability, are a long-term feature of the market even when the scheme is wound up. (Flood Re, 2016a, pp. 7-8;emphasis added) The company has also outlined the premises informing its market-privileging agenda; as befits a product of neoliberal Britain, these involve anticipated efficiency and consumer-welfare gains vis-à-vis the non-market conditions that will prevail during Flood Re's own lifetime, when, as we will shortly see, high-risk households will continue to rely on a cross subsidy from those with little risk of flooding. Flood Re also introduces operational costs that an open market would not incur. … [T]hese market distortions mean that an insurance market with riskreflective pricing for household flood risk would provide a more efficient solution and offer the potential for better outcomes for households at significant risk of flooding and those with low risks of flooding. (Flood Re, 2016a, p. 11) Recalling Collier's comments on the particular moral economy associated with market-privileging insurance schemes, Flood Re's long-term agenda is also clearly about individual 'responsibilization'. As the Tory MP Richard Benyon said in parliamentary debate on the proposed scheme: 'We should encourage households to see the process as a transition under which they will be rewarded when they take responsibility. If they take measures to reduce the flood risk to their property, they will benefit' (HC Deb 25 November 2013, c82).
Before turning to Flood Re's organizational and operational details, it is important to understand why exactly it was that Britain, in 2013, was not yet considered ready for the free market in flood insurance that all of those involved in Flood Re's developmentin industry and government alikeagreed was ultimately desirable. Why, in Gramsci's terms, could the new not yet be born out of the ashes of the old? To answer this question it is necessary to pause briefly to consider neoliberal explanations for situations where markets fail to deliver the idealized efficiency and welfare outcomes. The failure in such instances, it is claimed, is not of markets per se. Rather it is typically a failure, firstly, of states to 'do' markets properly (perhaps being too 'interventionist'), and/or, secondly, of market actors of various kinds to behave in the ways that markets require (i.e. 'rationally'). It is this second explanation of failure that is relevant here. According to neoliberal orthodoxy, functioning markets require rational market actors. As Christian Berndt (2015) has shown, the much-lauded sub-discipline of behavioural economics has played a pivotal role in the formulation of this orthodoxy, doing so by shifting attention 'from the market to the market subject, that is, from market failure to behavioural failure' (p. 569).
This is how Britain's non-readiness for a free market in flood insurance at the time of the expiry of the Statement of Principles was essentially articulated: as a case of behavioural failure. Key actors simply were not behaving in the 'right', rational way; or at least, they were not yet considered ready to do so. The principal failure, it was felt, was of high-risk households themselves. Thus Flood Re has bemoaned the fact that steps to increase property-level resilience have historically only been taken by around a third of previously-flooded households and by just 6 per cent of those who are aware of their flood risk but have not been flooded, despite the government offering to partially or wholly fund the necessary interventions (Flood Re, 2016a, p. 34). Meanwhile, the ABI has remarked that, at the present moment in time, house-hunters are more likely to ask about parking provision than flooding (Milligan, 2016)a claim supported by recent survey research showing that more than half of UK homeowners have never checked whether their homes are in at-risk areas (YouGov, 2017). Notably, Flood Re (2016a, pp. 34-35) explicitly invokes behavioural economics in seeking to account for such 'irrationality' (c.f. Harrabin (2016), for an alternative set of explanations).
But it was not only households that were seen to be behaving in ways unconducive to a free market. Local authorities were, too. Specifically, they were failing to curtail further flood-plain development, despite the fact that properties built after 1 January 2009 were not covered by the revised Statement of Principles. In other words, councils were failing to respond 'rationally' to the 'signal' that 2008's revision of the Statement had theoretically sent them. As recently as 2015, new houses were still being built in England's highest-risk flood areas at almost twice the rate outside flood plains (Gosden, 2015).
Could a free market be safely introduced in the context of this type of behaviour? The government thought not.
And, at least according to the political opposition, the government itself was not behaving propitiously, either. In parliament in late 2013, the Labour MP Maria Eagle castigated 'severe cuts to investment in flood defences' (HC 25 November 2013, c65). Such behaviour not only contravened the spirit of the historic pact between the government and the insurance industry. It also militated against the realization of a market for flood insurance that would be affordable for all households, and thus one where the government was not required to intervene either to enforce inter-household cross-subsidization (as historically had been the case) or as insurer of last resort.
Britain's non-readiness, in other words, was multifaceted and manifest, and translated into what Roger Harrabin (2016) has described as 'a complex blame game over responsibility for floods': Hundreds of thousands of householders in flood risk areas have failed to install basic protection against rising waters, insurers say. … The insurers have been criticised by the Environment Agency for failing to protect inundated properties. … Local councils are also part of the meleethey want more cash for flood funding from the government, and more control of how it is spent. They are critical of the Environment Agency.
Local authorities blame the government (in the shape of the EA); the government blames the insurance industry; the insurance industry blames homeowners. And homeowners? In Tewkesbury, Meek (2008) found a surprising equanimity towards the private sector, alongside 'consistent venom' towards its public counterpart: there is more hostility towards the government, the council and the Environment Agency for not stopping housebuilders than there is towards housebuilders for building houses … When insurers raise their premiums, more blame is directed at the government for not spending enough on flood defences than at insurers for raising the premiums.
Whatever the vector of blame, the broader point about Britain in the early 2010s is this: Nobody, in short, appeared ready for the market.
This, then, was the febrile context in which the government and the insurance industry sat down to endeavour to hash out what might come after the termination of the Statement of Principles in 2013. The first meetings, in the form of a 'Flood Summit', took place in September 2010. But, as we have seen, as late as March 2013 no agreement had been reached, despite the fact that the Statement only ran until the end of June. In the end, a provisional agreement was reached at the eleventh hour. The basis for this agreement was the industry's Flood Re proposal; and although Flood Re in practice, as launched in 2016, is not exactly the Flood Re as scoped in that proposal, the core ingredients, which we will turn to presently, were already there. It took the best part of another three years, however, for Flood Re to become a reality. It required new government powers, which were introduced through the Water Act of 2014 and came into force in January 2015. And it required subsequent regulation -The Flood Reinsurance (Scheme and Scheme Administrator Designation) Regulations 2015 and The Flood Reinsurance (Scheme Funding and Administration) Regulations 2015to spell out the detailed rules of the scheme (which were debated in parliament through early 2015) and thus to enable the exercise of those new powers. As all of this was being worked through, the coverage of the Statement of Principles was repeatedly extended, ultimately until Flood Re came into force on 4 April 2016.
Flood Re is essentially an industry-owned reinsurance vehicle, designed in the first instance, as noted earlier, to make flood insurance available and affordable to (most) high-risk households (there are exclusions, which I will address shortly). Households buying home insurance do not deal with Flood Re directly. They continue to buy home insurance from insurers as usual, and the latter continue to have the freedom to set individual premium and excess terms. But insurers now have the ability (though not obligation) to pass on specifically the flood risk of each policy to Flood Re. They can do so by buying reinsurance for that risk so that, in the event of a claim, Flood Re reimburses the insurer for the flood element, with a standard deductible of £250. This reinsurance costs the insurer an amount (the Premium Threshold) based on the council tax band associated with each property. These thresholds have been set sufficiently high that insurers have no incentive to cede properties at low flood risk, but, crucially, 'below the level that would be charged for properties with the highest risk if prices fully reflected those risks'. 'This', Flood Re (2016a, p. 22) explains, 'provides a subsidy for those properties judged to be at risk and improves affordability'.
In other words, because actual risk is not being fully priced, the expectation is that Flood Re will pay out in reinsurance claims more than it receives in reinsurance premiums. This is the subsidy that it refers to. How is this subsidy being financed? In the first instance, by the industry itself: a levy (Levy 1) is being raised from all insurers writing home insurance in the United Kingdom. Set by an independent actuary, this levy is payable proportionate to market share, totals £180 million per annum for Flood Re's first five years, and will be reviewed every five years of Flood Re's 25-year life. (If Levy 1 is not sufficient in any given year, Flood Re can also issue a compulsory call for additional funding from the industry through a further levy, Levy 2; the government meanwhile remains primarily responsible for any losses due to a 'catastrophic'more than 1:200 year lossevent that Flood Re cannot meet, as well as for increased spending on flood defences and for limiting risky flood-plain development.) But while the levy mechanism provides the immediate funding of the subsidy, the latter is ultimately being met by low-risk households, just as it was under the Gentlemen's Agreement and Statement of Principles. For the industry is fully expected to pass on the cost of the levy to its wider customer base (Ralph, 2016a). In this sense, and as Defra (2013c, p. 1) explicitly acknowledged when initial agreement on Flood Re was reached, the scheme ultimately serves to 'formalize the existing cross-subsidy between low and high flood risk policyholders within the market'. At the time of launch, it was expected that up to 350,000 homes could benefit from the scheme (Flood Re, 2016b, p. 8). A year later, Flood Re (2017, pp. 8-9) reported that the proportion of households with previous flood claims able to get quotes from two or more insurers had increased from 9 per cent prior to launch to 95 per cent, and that four out of five such households had seen a reduction in the price of available quotes of more than 50 per cent.
The main line of criticism of Flood Re has concerned its exclusions. These include all properties built since the start of 2009, as was the case under the revised Statement of Principles. But exclusions also include all commercial property, buy-to-let residential properties (where buildings insurance is considered to be commercial), purpose-built blocks of flats (for the same reason), and flats in converted houses where there are four flats or more. In total, according to the British Property Federation, there are some 800,000 properties in Britain that are at risk of flooding but nonetheless ineligible for the scheme, of which 70,000 are high-risk (Palmer, 2015). Criticism has also been made of Flood Re's inclusions. It covers top-end riverside mansions in council tax Band H, even though the original proposal was for these to be excluded (Defra, 2013c); indeed, it is expected that homeowners in affluent but relatively flood-prone parts of the south-east of England such as the Thames Valley will be Flood Re's 'biggest beneficiaries' (Ralph, 2016b).
My own critique of Flood Re, or rather of the government that legislated it into being, is different. Fleshed out in the final section of the paper, it is concerned less with Flood Re's short-term objective of ensuring the continuing availability and affordability of flood insurance, than with its long-term objective of ensuring a smooth transition to a free market. As I have emphasized from the outset, Flood Re is understood by its architects in industry and government as a means to an end as much as an end in itself. It is, perhaps even primarily, a vehicle designed to get Britain to a functioning market scenario. Doing so is in fact no less than its statutory responsibility: to 'manage, over the period of operation of the scheme, the transition to risk-reflective pricing of flood insurance for household premises', as paragraph 64 of the Water Act (2014) stipulates it.
How then is Flood Re expected to do this? To understand the expectations vested in Flood Re in this regard, we need to return to Christian Berndt and behavioural economics. As Berndt (2015) shows, behavioural economists did not throw in the towel when they discovered that market actors often fail to behave in the 'rational' manner that efficient markets require of them. On the contrary, it is a central premise of (some) behavioural economics that 'rational behaviour can be learnt' (p. 568): the analytical shift from market failure to behavioural failure associated with behavioural economics was, at once, and more 'positively', a shift from 'market regulation to behavioural engineering' (p. 569). In other words, the ultimate challenge that behavioural economics sets for itself is to engineer market-conforming behaviours. Its answer to this challenge lies in incentivization: 'the construction and management of incentive structures in order to channel the behaviour of "humans" into a direction that is deemed socially beneficial' (p. 572), or so-called 'nudging'. Berndt explores this approach in the context of anti-poverty policy in the Global South aimed at 'aligning individual behaviour with idealized notions of market rationality' (p. 568).
We can, I suggest, read Flood Re's long-term remit in the exact same way. Sure, in 2013, Britain was not considered ready for a free market in flood insurance, because the evidence suggested that key actors could not be trusted to behave appropriately. But Britain could be made ready; those actors could be encouraged or educated, if you like, to behave rationallyif the incentives were gotten right. Indeed, the fact that, in the past, the incentives had not been gotten right was used as an argument to explain Britain'sits key actors'non-readiness: the country had reached the impasse at which it found itself in 2013, Maria Eagle suggested, precisely because of the 'failure of the Government's proposals to strengthen incentives for the uptake of household flood protection measures' (HC Deb 25 November 2013, c66). The primary task facing Flood Re, therefore, was to put 'the right incentives for public bodies, businesses, communities and individuals to manage flood risk' in place (Defra, 2013b, p. 13). The transition to more risk-reflective pricesthat is, to a free marketshould be based, Defra said, 'on robust evidence of local risk, to increase the incentives for flood risk to be managed whilst allowing time for choices to be made and appropriate action to be taken' (Defra, 2013b, p. 17). Understanding this approach helps explain the specific lifetime allocated to Flood Re before it would be wound up. How long would it take for public bodies, businesses, communities and individuals to adapt to the incentives introduced by Flood Reto, in effect, fully respond to their behavioural therapy? 'The Government envisages this transition taking place over the next 20-25 years' (Defra, 2013b, p. 17). Hence, Flood Re would live for a quarter of a century; it was, said Tory MP Thérèse Coffey, 'designed with a 25-year lifetime to help householders at high flood risk to adapt to risk-reflective pricing' (HC Deb 6 December 2016, c54WH).
Needless to say, the process of behavioural adjustment would be gradual; the expectation was not that actors unaccustomed to managing their risks 'responsibly' would suddenly begin doing so after 25 years. They would learn, and respond, progressively. Thus, the other component of the planned transition would be a commensurate gradual removal of the crosssubsidies instituted by Flood Re. Not only would these subsidies in theory become less and less necessary as actors became more and more accustomed to behaving 'rationally'; removing the subsidies would itself encourage the process of adjustment, adding an element of stick to the carrot of lower insurance premiums. Hence the company's notification that 'between now and 2039, Flood Re will seek to reduce the subsidy provided through the scheme and the associated industry levy' (Flood Re, 2016a, p. 11). Weaning dependents off subsidies while/by incentivizing responsibilization: the very quintessence, one might say, of paternalistic, interventionist neoliberalism. In any event, the consolidated vision embedded in Flood Re and to be delivered by it is of a British flood insurance landscape where, by 2039, subsidies have been reduced to zero and all actors are suitably disciplined to act rationallya vision of, in short, a free market.
What more can we say about the market that will, in theory, succeed Flood Re, other than the notion that it will be 'free'? Very little. If one curiosity of Flood Re is that it was conceived and introduced in order to both avoid (in the short term) and seed (in the long term) a free market in flood insurance, another is that this 'free market' remains stubbornly ill-defined. In fact all we are told about it is that it will be 'free', in the sense that pricing will reflect risk. But this says nothing about availability, nor, as I shall discuss in the next section, about affordabilitywhat, in practice, are 'risk-reflective' prices? None of the parties to Flood Re's establishment and operation, in government or industry, has deigned to fill in these blanks. The 'free market' is simply out there, on the horizon, a holy grail to be quested for but which in reality is inaccessible cognitively as well, for now at least, as practically. If this haziness sounds improbablewhich it well might given the time, money and energy which has been and is being invested in the pursuit of this visionthen we need look no further than Flood Re itself for substantiation of the fundamentally nebulous nature of the very thing it is supposed to be working towards. The company knows its mission is risk-reflective pricing, and that it is statutorily bound to deliver. And yet it concedes: 'The statutes setting up Flood Re do not explain what risk-reflective pricing means in practice' (Flood Re, 2016a, p. 6). Indeed they do not. The 'market' to be actualized is, instead, merely symbolic, figurative. Gestured to rather than actively delineated, it is, in sum, allusive.

Morbidity
It is, needless to say, much too early to say with any certainty what the long-term effects of Flood Re will be, but, in all probability, it will fail to effect a successful transitionsmooth or otherwiseto a free market in flood insurance in which high-risk households can access affordable policies. For one thing, historical experience teaches us to be sceptical. Upon agreeing the final revision to the Statement of Principles in 2008, the then Labour government confidently stated its beliefin terms to be echoed with uncanny exactitude eight years later upon the launch of Flood Rethat a functioning market, if not the explicit objective on that occasion, would be in place by mid-2013 and would 'provide flood insurance to the vast majority of households and small businesses efficiently and without the specific commitments [contained in the Statement]' (Defra, 2008). We know how that turned out. And geographical experience teaches us the same lesson. As the Tory peer Mohamed Iltaf Sheikh observed in a 2014 Lords debate (HL Deb 27 January 2014, c1050) on the Flood Re proposals: The ultimate goal of Flood Re is to pave the way for a smooth transition to the free market. If that were successful, it would be most welcome, but, of course, that is not an easy feat. No country in the world has a free market for flood insurance which successfully preserves widely available and affordable flood insurance for those at high flood risk without some form of government involvement.
Nowhere else in the world has achieved it, so why should we expect Britain to be different?
For there are, of course, very good reasons why no country in the world has achieved the type of market to which Flood Re's creators consistently allude. Not the least of these is the simple fact that some properties in some locations are of such high risk that no amount of 'nudging' is going to make much of a difference. In the context of Britain and Flood Re, nobody has made this case more clearly and passionately than the Labour MP for flood-prone Hull, Diana Johnson. In a November 2013 Commons debate (HC Deb 25 November 2013, cc82-95), Johnson explained to MPs unfamiliar with her constituency that 'in areas such as mine, Hull, where 90 per cent of the city is below sea level, home owners and home builders can do all they can, but we will always be at risk of flooding'. It's just the way it is. So while she approved of the subsidy mechanism to be introduced by Flood Re, she was deeply concerned about what would happen in the longer term. I hope that the scheme that we end up with will not just disappear after the 25 years. … If the free market is just opened up, we will be left with no insurance companies that want to offer insurance in those areas, because the risks are too high.
The legal scholar Mateusz Bek (2015, p. 39) makes a comparable, equally blunt, point about the basic socio-environmental realities faced (but not faced up to) by Flood Re. 'It is', he writes, 'unlikely that relatively poor homeowners living in areas prone to flooding will ever be able to afford market rates for flood cover'.
It does not help matters, moreover, that certain aspects of Flood Re's design and mandate explicitly militate against its achievement of an objective that realistic commentators agree is difficult enough (if not impossible) to begin with. Three such aspects, connected to one another in various ways, merit particular attention. The first concerns understanding and management of flood riskand thus its costs, insured or notper se. Flood Re says (2016a, p. 8) that for the transition to a free market with widely available and affordable insurance to happen, 'the risks of flooding need to be managed and the cost of insurance claims reduced'which is surely true. Yet, as constituted, Flood Re is fundamentally disconnected, in more ways than one, from the reality of those risks.
A steady-state view of a deeply unsteady world was woven into it at launch. As Labour's Barry Gardiner noted in late 2015, the scheme 'does not address the rising cost of flood risk [projected] by the Committee on Climate Change and the Environment Agency' (HC Deb 15 September 2015, c6)his party colleague Maria Eagle had previously lambasted 'the folly of ignoring the impact of climate change in the Flood Re insurance scheme' (HC Deb 26 February 2014, c325). One recent estimate has annual UK flood damage costs rising from £1.1 billion today to £27 billion by 2080 (Burton, 2017). Just because existing flood defences are deemed adequate today, providing the EA with comfort in approving flood plain development, does not mean they will be adequate in the future. When, in late 2015, some such defences were overtopped by 'unprecedented' floods, the head of the Committee on Climate Change's adaptation sub-committee, the influential zoologist Lord Krebs, commented: 'Today's unprecedented may be tomorrow's norm ' (cited in Gosden, 2015).
Flood Re's structural indifference to the rising costs of flood risk in a warming world would perhaps be less problematic if it were designed in such a way as to contribute to reducing those costs. But, for all the talk of incentivization, it is not. In its December 2013 commentary on the Flood Re proposals, Krebs's sub-committee noted that Flood Re would actually largely remove any financial incentive for flood risk to be managed amongst householders covered by the scheme [because] existing price signals will be largely removed. High risk households will pay the same for their insurance no matter how severe their risk, and whether or not they take steps to protect themselves. (CCC, 2013;emphasis added) The environmental economists Surminski and Eldridge (2017) judged the proposals similarly, saying Flood Re had only limited potential for risk reduction because pricing signals and incentives were overridden; there were 'no requirements for built-in incentive mechanisms' (p. 421), such as lower premiums for households adopting risk reduction measures. Notably, a number of proposed amendments to what became the 2014 Water Act were tabled that would have required Flood Re to both factor in projected escalation in flood risk and associated costs and to incentivize owners to invest in flood protection measures in the types of ways suggested by Surminski and Eldridge. But the Tory-led coalition government voted those amendments down (HC Deb 26 February 2014, c324;HC Deb 15 September 2015, c6).
And these were not the only amendments it declined to supportwhich brings us to the second aspect of Flood Re's design that militates against the achievement of its objectives. Critics of the initial Flood Re proposals expressed concern that key actors in the flood insurance landscape, households in particular, would be unlikely to be incentivized to change their behaviour if they did not know that Flood Re was only a temporary solution and that, over the course of its lifetime, premiums would be adjusted to increasingly reflect risk as subsidies were gradually removed. Markets can only possibly work, it was suggested, if relevant information is freely availableand, as initially scoped, there was no guarantee that under the Flood Re scheme it would be. Such critics therefore proposed modifications specifically to ensure that householders would be given the relevant information. Here is Lord Krebs, a key advocate of this, in early 2014: Flood Re has been designed to be invisible to the households concerned. As currently cast it will not give them any source of information. It is important that we … alert householders who might be affected to the fact that Flood Re is not a permanent arrangement, and that there will be a transition. … [I]t is important that householders over that 25-year period take action if they are at risk, to reduce their risk. Therefore it is important for households to have transparent information about the nature of the transition to risk-reflective prices that will arise at the end of Flood Re. It is hard to object to giving consumers information. (HL Deb 11 February 2014, c586) But the government refused to back amendments to this end, too (HC Deb 15 September 2015, c6). The relevant section of the final 2015 regulations (UK Government, 2015, para. 26) states that Flood Re 'must provide' relevant information each year to insurers using the scheme, 'for the purposes of enabling those insurers to supply that information to holders of those policies'. But only the former part of that information chainfrom Flood Re to the insurers is governed by compulsion ('must provide'); the latter, from insurers to consumers, is left open to insurer discretion ('for the purposes of enabling'). The third and final aspect of Flood Re's constitution that makes what is by nature a difficult challenge even harder is the markedly limited nature of its powers. As we have seen, Flood Re has a statutory responsibility to manage the transition to a free market. But as the company itself notes (Flood Re, 2016a), it is others that will have to change behaviours to ensure the transition takes place: 'This will require action from Central and Local Government, including the Devolved Administrations, insurers, households and a number of other parties' (p. 8). Perhaps the key paragraph in all the documents thus far produced by Flood Re is this one: it is clear that the majority of the actions that will need to be taken to reduce the costs of providing insurance are not under the direct influence of Flood Re. For instance, Flood Re does not control decisions over flood defence investment and maintenance; insurers' pricing; the behaviour of households; building regulations and choice of location for new development; or policy on climate change and adaptation. This means that, alongside Flood Re's approach to transition, others will need to take action if an affordable market is to be achieved. (Flood Re, 2016a, p. 13;emphasis added) All Flood Re really has the power to do, as it says, is to 'support and promote ' (p. 8). No wonder then that the three work programmes for 'the coming months and years' laid out in the first of the five-yearly transition plans Flood Re is required to produce -(1) benchmarking and tracking prices and availability of household flood insurance, (2) understanding how its data might be shared to help other actors make decisions over the action needed to manage the risks and costs of flooding, and (3) providing the evidence base needed to understand how households and insurers can be supported to manage the risks of flooding and reduce the costs of claimsare, to put it mildly, underwhelming (Flood Re, 2016a). They could scarcely be otherwise.
There is, in short, a fundamental mismatch between Flood Re's powers on the one hand and its responsibilities on the other. The government has simply not given Flood Re the tools it would need to have any chance of getting the long-term job done. 'The fundamental tools to create an environment for risk reflective pricing [such as prohibition of inappropriate floodplain development and a substantial increase in funding for investment in flood defences] lie ', as Bek (2015, p. 38) says, 'with the government and not with Flood Re'.
For all the above reasons, it is impossible therefore to disagree with Bek's judgement that, on the basis of Flood Re as currently constituted, a successful transition to risk-reflective pricing is 'inconceivable'. Success, Bek argues, would require a radical change in certain policies. … Even the ABI recognised that in the context of climate change, a major shift in approach and resourcing over time will be required in order to achieve affordable flood cover at the end of the life of the scheme. (Bek, 2015, p. 38) Among those who concur with Bek are the (re)insurance legal specialists Carter Perry Bailey, who doubt both whether households will 'have the incentive fully to flood-proof their properties' and whether the government will 'improve the flood defences to the necessary extent' (CPB, 2016, p. 3). Similarly, the Institute and Faculty of Actuaries, which concluded that 'as long as high flood risk persists, an affordable market based on risk-reflective pricing in unlikely to be achieved' (IFoA, 2016, p. 1). In fact even Flood Re itself appears to have (well-founded) doubts. In its first transition plan, as if making its excuses for failure ahead of time, it repeatedly tells readers that the 2015 regulations that established the company actually make no specific stipulations regarding the affordability of flood insurance once the fêted 'risk-reflective market' is in place: 'It could simply mean returning to the unaffordable premiums and excesses we see today' (Flood Re, 2016a, p. 6). Such an outcome is clearly not the company's intention; but, behind closed doors, it might well be its expectation.
No less interesting and important than the question of whether a tolerable free market is what comes after Flood Re is the related question of what the implications are of the government presuming and saying that it will beand of alluding confidently to that imagined, harmoniously marketized, future.
The answer to this latter question is that Flood Re, and the transition to the market that it is expected to effect, essentially permits the government to rest up. Offloading the problems of managing the growing financial risks of flooding to a mythical market gives the government permission to not itself have to confront those problems today in a substantive, sustainable and no doubt extremely expensive way. In this sense, Barry Gardiner is right to say that the government has not so much given responsibility (without power) to Flood Re; it has abdicated responsibility (HC 27 October 2015, c7). Abdication is indeed the fundamental work that the trope of the allusive market performs. The allusive market is ultimately an exceptionally useful crutch, and all the more useful and powerful for the fact that it will not be until 2039and of course a different government administrationthat judgment will be made as to whether this emperor has any clothes. As the climate finance scholar Aled Jones has observed, Flood Re absolves the government of having to make difficult decisions now, such as whether, when and how to strategically withdraw from risky assets (see Purcell, 2017). Itand the allusive market it promisesis, as Gardiner told parliament in no uncertain terms, a way of avoiding the problem and any need to make the uncomfortable acknowledgment that the price that reflects increased risk will be greater where the risk of flood becomes greater as a result of lower Government investment in flood defences, increased building on the flood plain or adverse climate change. That is the fear. (HC 27 October 2015, c8) Don't worry, the government says; the market will save the day. But in reality, and precisely because of this promise, what the allusive market does is simply defer the day of reckoning, which is to say the day when the tough decisions need to be made and the difficult actions carried out. Flood Re metaphorically kicks the can down the road. Bek (2015, p. 39) is quite right to worry that 'we will be back to square one on 14 May 2039'. Or as Carter Perry Bailey's lawyers put it: 'The suspicion must be that in 2039 another short to medium term fix will be required. Rien ne dure comme le provisoire' (CPB, 2016, p. 3).
In the meantime, with hopes and expectations firmly vested in the market and with the government released thereby from the obligation to grasp the nettle itself, morbid symptoms, unsurprisingly, proliferate. And therein lies the rub. In deferring the day of reckoning, the reckoning almost certainly becomes more daunting. It is in fact improbable that, in 2039, Britain will merely be back to 'square one'. Precisely because a meaningful long-term strategy to address the underlying problems of flood risk management has not been developed now, those problems will continue to mount. In 2039, the necessary decisions will be tougher, and the necessary actions even more difficult. So back, more like, to square minus-one.
The particular morbid symptoms highlighted by Lord Krebs are the continued draining of peat bogs on Britain's uplands and the continued spread of concrete and paving tiles over gardens and other green spaces (Gosden, 2015). The former trend, for which farms and grouse shooting estates bear particular responsibility, worsens flood risk by reducing upland bogs' capacity to act as a natural 'sponge' for rainfall; the latter does the same by preventing water draining naturally. Here the main concern is flash floods, where heavy rain overwhelms sewers, and which represent the most frequent type of flooding in Britain. At present, new housing developments can simply connect to existing drains, which increases the risk of floods. As Damian Carrington (2017) has reported, a law requiring new developments to incorporate sustainable drainage systems (SuDS) like ponds, green roofs and permeable pavingwhich can help to slow the flow of water into sewer systems, hence cutting flash-flood riskwas passed in 2010, but the government promptly 'put the rules on hold, aiming to save developers money and speed up housebuilding'; and in 2016 it 'rejected proposals to use its Housing and Planning Act to increase the use of SuDS'. (Krebs's reaction? 'In 20 years' time, people will look back and say, "What were they thinking?"') The cumulative result of these and other ongoing trends deplored by flood-risk experts is that, already, according to a recent study by geomorphologists at the University of Salford that received coverage in the national media, as many as around nine-tenths of England's extensive flood plains have substantially lost the ability to effectively serve their historic function of 'soaking up excess water in their vegetation, forming natural buffers that hold back or divert rushing water after rain, and providing areas where rivers can breach their banks and wetlands can be replenished' (Harvey, 2017). The study in question laid much of the blame for the damage to the nation's flood plains at the door of intensive agriculture, which, it suggested, 'has created artificially "smooth" and uniform landscapes, with hedgerows removed, large areas given over to single crops, wetlands drained and woods and grassland diminished' (Harvey, 2017).
But building on flood plains also 'has been singled out for years as a key problem' (Harvey, 2017). The National Planning Policy Framework requires local authorities to avoid flood risk areas 'wherever possible. Where building in these areas is being considered', Kirby (2017a) explains, 'councils must ensure that suitable alternative sites at lower flood risk are not available, that new developments will be resistant to flooding and safe for its lifetime, and flood risk will not be increased elsewhere'. Councils also have to pay heed to the views of the EA. But although the EA advises on all proposals for major development in areas of medium or high flood risk, local authorities, as an EA spokesperson points out, 'make the final decisions' (cited in Kirby, 2017b). Since the coalition government relaxed planning rules in 2011, furthermore, it has become easier for councils to approve planning applications in high-risk areas, and they are no longer required to report cases where they overrule EA advice (Chelmi, 2016).
Research based on a freedom of information request found that in the five years from 2011/2012, councils approved more than 4,400 residential units in spite of EA objections about flood risk, and that more than half of these were in National Flood Zone 3 (the highest level of risk) (Kirby, 2017b). In fact, the proportion of all new residential addresses in England created in Zone 3 steadily increased during the years of Flood Re's scoping and establishment, from 7 per cent in 2013/2014, to 8 per cent in 2014/2015, and 9 per cent in 2015/2016. 2 And subsequent to Flood Re's launch? It leapt still further, to an historic-high of 11 per cent in 2016/2017 (MHCLG, 2018, p. 7). So much thus far, then, for incentivization of 'market rationality'. The very concept would, indeed, be laughable in some parts of the country. In early 2017, the indefatigable Diana Johnson asked the government to publish regional data on the aforementioned metric, which showed that between 2013 and 2016, six English local authority districts saw more than half of new residential addresses being created in Zone 3. The two highest percentages were 90 per cent, in Johnson's own town of Hull, and, staggeringly, 98 per cent, in Boston in Lincolnshire (Ottewell, 2017). It does not get much more morbid than that.
But none of this should necessarily be surprising. There are no more influential lobby groups in Britain than the property sector and the financial sector that funds its land acquisition and development strategies. Development patterns reflect its appetites, and the apparatus of government tends to acquiesce. Johnson herself was probably not surprised by the data. In late 2013, she had expressed incredulity at finding that the government was itself promoting new development in an area of Hull (the Kingswood estate) within Flood Zone 3, when the entire point of excluding properties built after 1 January 2009 from both the Statement of Principles and Flood Re (then working its way through parliament) was supposed to be to discourage development in such high-risk areas: I am really concerned that when I asked the Secretary of State for Communities and Local Government in the House this afternoon about what discussions he has had with his colleagues in the Department for Environment, Food and Rural Affairs, he did not seem to know that there was a problem. Yet his Department is at the moment promoting the Help to Buy scheme extremely heavily in Kingswood to get people to buy new homes in an area that another Department says will be completely excluded from the Flood Re scheme. There is a misselling issue there, with people not being fully aware of what this Government are doing. (HC Deb 25 November 2013, c94) Ultimately, against the backdrop of a national socio-environmental landscape increasingly being ridden roughshod over by what Mike Davis (2017), in the US context, calls 'the financial and real-estate juggernaut', Flood Relegitimated by the allusive market that is slated to succeed itamounts to little more than fiddling while Rome burns. Or sinks.

Conclusion
The bulk of the critical heterodox literature on markets and neoliberalism has focused, for good reason, on how the former have become the principal organizing infrastructure of society and economy in countries dominated by the latter, and on the often problematic consequences of this marketization of socioeconomic life. Much less has been written, however, about the role of what we might call neoliberalism's market 'undertakings'. While scholars have certainly examined the performative effects of the active deployment of market models, and the (in many ways no less performative) effects of attempts to justify various forms of social action and worth by means of reference to market logics, they have not explicitly considered the implications of policies that hinge on the premise that there will be a functioning marketonly, not just yet. This paper has examined an example of one such undertaking: the UK government's pledge that by the time Flood Re, its new, subsidy-based scheme for insurance of flood risk, is wound up in 2039, a free market for such insurance that works for all households, including those at highest risk, will have materialized. The paper has shown that while Flood Re does not itself provide Britain with a market, the premise that a market will succeed it is central to its constitution and, indeed, its very legitimacy; it is unlikely that the insurance industry, or indeed the government's own neoliberal firebrands, would have tolerated a 'solution' to the management of the financial risks of flooding that was not at least bookended by a promissory market arrangement. But the paper has also shown that the likelihood of this allusive market ever becoming reality is exceedingly slim, not least given the actual design of Flood Re, which in many crucial respects militates against the feasibility of transition from subsidy to functional and affordable market 'freedom'. In this regard, one might argue that Flood Re and the wider policy stance it embodies evinces not so much what Judith Butler (2010) calls a 'market presumption'the presumption that a market is possible as a market pretence. There is simply too much counter-evidence, and too many countervailing forces, for a satisfactory market eventuality to be a reasonable expectation.
This matters because the interregnum associated with the inability of the market to be born is also characterized by a refusal of the state to countenance a viable long-term non-market solution. This, in fact, is the principal effect of the market allusion: to eliminate the need to envision and effect an alternative approach, on the grounds that the market is coming and will ultimately suffice. As the paper has shown, this interregnum provides fertile soil for the germination of all manner of morbid symptoms that a more robust approach to the management of flood risk insurance that does not rely on the crutch of an imagined future market utopia might have the capacity to help forestall. Given that there appears to be little prospect of action being taken to remedy this morbidity while the allusive market retains its cognitive grip and policy hegemony, it is perhaps not unwarranted to suggest that pretending that a perfect market is not just possible but portending, when it almost certainly is not, is at least as dangerous as successfully putting an imperfect market in place.