Managed retreat and planned retreat: a systematic literature review

Abstract

Vulnerable locations, such as coastlines, are at a high risk of loss and damage. Such places will suffer deleterious impacts as climate change impacts are increasingly realized. As societies try to adapt to these impacts, managed or planned retreat—aimed at moving people and assets away from vulnerable locations—is gaining increased attention. Despite this increased attention, systematic literature reviews of the retreat literature remain scarce. This paper undertakes such review and uncovers a marked increase in retreat scholarly research papers in the past 5 years. An analysis of 135 managed and planned retreat journal articles is completed. Findings include a strong emphasis on regional or local case studies exploring governance, policy or institutional settings and levers across a range of geographies. Property rights and market interventions, such as compensation schemes, evidence the prevalence of neoliberal predilections. This emphasizes the importance of renewed engagement with political economy scholarship vis-à-vis climate change adaptation, also supported by the sharp increase in the evidenced social and environmental justice impacts.

This article is part of the theme issue ‘Nurturing resilient marine ecosystems’.

1. Introduction

Coastal locations and ecosystems are highly prized, highly valued places around the world. They are prized and valued for their economic and environmental benefits, and for their social and cultural significance [1,2]. Because of this, they are often also highly developed locations, with human populations and supporting infrastructure around the world concentrated on or near coastlines. This growth is intensifying even as both climate change impacts and past and current human incursions on coastlines erode and degrade coastal ecosystems. The concomitant desires for coastal living [3–5] combined with now rapid climate risk and biodiversity loss [6] mean that societies around the world are facing a stark contest: what to do with at-risk built environment, the people who own it, the institutions that govern it, and the people and ecosystems who reside in vulnerable locations?

Climate change adaptation literature and the Intergovernmental Panel on Climate Change (IPCC) have long recognized the environmental risks facing coastal locations and in doing so have concerned themselves with how to plan for such risks (e.g. [3,5–18]). Acknowledging that adaptive pathways planning for managed (and planned) retreat must be dynamic and iterative to be useful and responsive [15,19–22], coastal adaptation has frequently and historically been framed as requiring ‘protect, accommodate, retreat’ options for developed coastlines. This manipulation of the dynamic coastline is often positioned as a way to balance the varied and often competing economic, social, cultural, legal and environmental needs and interests in coasts [3,5,10,17,18,23–29]. Implementing these options is of course problematized by climate change impacts such as storm surge, sea-level rise, saltwater intrusion, coastal flooding and biodiversity loss, each occurring against that backdrop of these diverse, often diametrically opposed, interests.

Coastal protection can take the form of soft or hard engineering measures, so named for their impact on the natural ecosystem. These are designed to be implemented prior, or in response to, the anticipated negative impact on the coastline. Note that negative impacts on coasts including their ecosystems can occur as a result of, for example, ill-placed and/or ill-designed development [30]. Soft protection measures may include things like sand dune renourishment, intended to maintain a particular shape or geomorphology of a coastline/beach. Sand dune re-nourishment is a favoured protection option as it can respond to sudden weather events or disasters, such as coastal erosion during coastal storms, by acting as a buffer to such events, minimizing the extensive loss of beach/coastline during an event. Hard protection measures include options such as permanent sea walls, often taking the form of concrete walls or groynes. These structures are permanently placed on coastlines to protect the assets landwards. Hard protection works last longer, but are more intrusive on the physical environment and on coastal ecosystems. This intrusivity includes making a coastline permanent in shape, meaning that the natural dynamism of a coastline is lost. This can have deleterious effects further along a coastline, where erosion can be significantly higher owing to the build structure [4,13,30]. On the other hand, and because coastlines are constantly moving owing to currents and other factors, coastal accommodation by for example raising floor levels of the built environment to accommodate water flows in flood or rising water table scenarios have become common. The third coastal development management option, managed or planned retreat, is more controversial. Retreat can therefore maximize the use of the coastal land over time by either allowing human settlement/development in some locations for a period of time, until the risks associated with the vulnerability of the location are too great at which time, the settlements or developments are lost, or relocated.

Not all, nor sometimes any, of these options for coastal climate change adaptation are viable. Many stagnate owing to political [14,15,22,23,31] and legal [3,13,15,17,18,29,32–35] tensions. When combined with varied and often conflicting social values [3,5] protect, accommodate and retreat options for climate adaptation and governance have been often been incremental, and where less incremental, certainly polarizing. Additional complications arise in Western countries, where inherited legal systems devalue Indigenous knowledge and prioritize institutional hierarchies that are less, or not at all, cognizant of environmental geographies, vulnerabilities and limitations [2,36–40].

So, with climate tipping points now reached [6,41] and vulnerable areas including coastlines at the increasing risk of loss and damage in many locations around the world, we must explicitly acknowledge that not all of these locations can be saved through protect or accommodate options. Retreat is a difficult, but a ‘must have’, in terms of design, process and conversations. Robust knowledge of the trade-offs and approaches in implementing managed or planned retreat is required to ensure that such retreat is undertaken in a way that nurtures ecosystems, is just and fair to the myriad of values attached to the location, and recognizes the highest and best use of the location at particular points in time [3,14,15]. It is also important to acknowledge the inherent difficulty of effectively implementing managed or planned retreat. We can learn from examples of attempting to implement managed retreat. For example, in New Zealand, a country largely focused on anticipatory climate adaptation, a large proportion of the land use planning mechanisms make provision for managed or planned retreat [42]. While managed retreat in New Zealand still remains difficult to implement, notwithstanding the legislative and institutional support for it [20,21], there is still much to learn.

To progress our understanding of managed and of planned retreat, and to move our understanding and recognition of both the peer-reviewed literature and the challenge of managed retreat forward, this paper offers a systematic literature review of published scientific literature on managed retreat and on planned retreat, for the period 1 January 2017–1 January 2022. Specifically, this paper is concerned to document the substance of research in this field and, in doing so, identifies both research gaps and future challenges. The paper provides analysis to measure managed and planned retreat papers, as evidenced in relevant scientific paper databases. It is relevant to note that the terms, ‘managed’ and ‘planned’ retreat are sometimes used interchangeably and so this paper has undertaken reviews of both fields. There are broader retreat terminology challenges, discussed in the following sections. Following the methods section, this paper describes and synthesizes the literature by theme, and then ends with discussion of the findings along with the identification of gaps and future research trajectories and challenges.

During 2021, other review articles concerned with managed retreat made their way into the literature. These undertook reviews by examining Google scholar papers on retreat as against emerging climate risk [43], or by undertaking a systematic academic database literature review examining preferences for coastal climate change adaptation [44]. The use and unique contribution of this review is that it is the first application of the systematic literature review method dedicated only to managed and planned retreat papers, using a systematic literature review method.

2. Methods

Systematic reviews are important mechanisms for how persons can survey and understand the current research and scientific literature in a particular field, on a particular topic, in a particular geography, or within a particular temporal scale or frame. Systematic literature reviews adopt features of data analytics in that there is a clearly defined protocol, which produces consistent and repeatable results through the use of the data inputs and formulae [45,46]. Of particular importance is the role of systematic reviews in aggregating and then analysing academic trends and knowledge development. Though this approach originated in the health sciences, it is now increasingly employed in the environmental and social sciences [33,44,47–49].

This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses approach [46,50,51]. A compiled list of publications was obtained through searches of Scopus and Web of Science for the time period between 1 January 2017–1 January 2022. These were imported into Excel in November 2021, with subsequent searches undertaken in late December 2021/early January 2022, to capture any 2022 publications and to ensure a full 5-year time period. The search strings we used to compile the publications were: TITLE-ABS-KEY (“managed retreat”) OR TITLE-ABS-KEY (“planned retreat”) AND PUBYEAR >= 2017–2022. In order to measure the growth of the field, additional AND PUBYEAR >= 2012 were completed for both search strings. The small number of earlier timeframe papers (i.e. 2012–2016) are not analysed as the purpose of this additional search is to show the growth of the field only. The search terms themselves were also necessarily focused on the use of double quotations around the entire phrases. Some limitations of this review include that literature may not be captured if it did not include the entire search term in the title, keywords or abstract.

Within these parameters, 110 Scopus papers and 120 Web of Science papers were returned for the search string ‘managed retreat’. Conversely, 18 Scopus papers (two duplicates, so removed) and 16 Web of Science papers were returned for the search string ‘planned retreat’. To avoid duplication, all managed retreat and planned retreat papers in Web of Science were read and reviewed, as these returned a slightly higher number and duplicated the Scopus returns. Owing to the (somewhat) manageable number of papers, all papers for managed retreat in Web of Science (119, after deducting one identified duplication), along with all the papers for planned retreat in Web of Science (16), have each been reviewed. Two of the 16 planned retreat papers were also contained in the managed retreat papers, so have not been reviewed twice. This process led to a total of 135 papers. These are organized thematically and textually analysed. Papers and themes arising are then summarized and analysed, in the results and discussion section. Other literature also appears in the introduction, discussion and conclusion, where it is directly relevant.

Expanding the review method adopted in this paper to include the year timeframe of 2012–2022, the managed retreat papers expanded to 142 in Web of Science (with a similar trajectory in Scopus). For the planned retreat papers, it expanded 26. This relatively small variation of papers as between 10 versus 5 years prior to 2022 demonstrates a marked increase in the subject matter over the past 5 years, academics' using managed or planned retreat terminology in papers, or both. My analysis of the returned papers found no material difference between the managed retreat and planned retreat papers, though a difference was anticipated. Both terms (and indeed other terms) are used interchangeably in discussing the moving of assets and/or people from vulnerable locations in response to a climatic threat. This review uncovered several different ways that authors were framing or describing managed or planned retreat (e.g. relocation, climate migration, resettlement, strategic retreat, pre-emptive retreat, anticipatory retreat, transformative adaptation and more). Doing this without a specific and explicitly identified scholarly purpose dilutes and confuses the field, and can fragment research effort. This interchangeability of broad term usage when referring to managed and/or planned retreat underscores not only the limitation of the search terms used in this review but—more importantly—the need for academics to be thoughtful in their composure of their journal articles, especially keywords. Academics would greatly assist the field by selecting the terminology that best describes the event/process/policy/other phenomena their retreat research is concerned with.

In many instances of this review, the climate/environmental risks or threats requiring retreat were coastally oriented (coastal storms, erosion and sea-level rise). Some of these were attempting to respond to these risks as they were occurring now, but most were exploring policy or perceptions relating to the anticipated future of coastal risks. The other commonly featured risk was flood, where papers explored retreat after the flood event. The overall trends and dominant themes across all papers are discussed below. No papers examined managed retreat relating other risk or threat types, for example with respect to bushfire or wildfires.

In the small number of papers returned using ‘planned retreat’ as the search string, there is some variance, in that only two papers were captured in both searches [52,53]. The additional 14 planned retreat papers are combined with the managed retreat papers, and all are discussed in the aggregate below.

3. Results and discussion

(a) . Trends

Many papers use case studies to explore the range of influencing factors relating to a geography, and/or key research questions to be explored. This is possibly owing, at least in part, to the relatively recent rapid growth of climate change adaptation scholarship outputs [54] and the increasingly interchangeable terms describing a detailed range of phenomena relevant to adaptation theory and practice [44]. The growth of climate change adaptation scholarship and the increasing recognition of the urgency of adaptive response including that managed retreat is also a factor [55,56]. Siders [57,58] hones in on this and explains that managed retreat must be scaleable in order to be viable; barriers to implementation must be identified and therefore better understood; practitioners must learn from historical events; and consistent policy approaches within nation-state jurisdictions are essential (on this latter point, see also [14]). Some returned papers make the case for specific disciplinary interventions in managed retreat, for example geology [59], sociology [60], and architecture and design [61] but the overall trend is one that is focused on the activity of retreat itself.

A number of the publications captured within this review are concerned with the breadth and range of terminology used to describe retreat [23,43,62]. This includes papers referring to managed retreat as anticipatory relocation, resettlement, relocation, strategic retreat, climate migration and more. Carey [23] posits that while the concept of managed retreat is not new, the addition of climate change threats as an impetus for managed retreat has meant that the term has become ‘politically perilous'. Their suggested remedy is to change the terminology, and instead of managed retreat, talk about strategic relocation or some other namesake. Dundon & Abkowitz [43] in their review paper also discuss at length the terminology issue, suggesting a range of different terms including their preference for ‘transformative adaptation’ so as to better encompass the positive aspects of retreat. Ajibade et al. [63] provide a comprehensive analysis of the differences between climate migration and managed retreat, noting that these are not synonymous terms (contra [64], who tag their paper as managed retreat, but predominately analyse climate migration presenting a tripartite typology: places that lose people, places that receive people and places that are optimal destinations). Ajibade [53] also provides a detailed analysis of the legal, political, economic and cultural aspects of retreat, which are also strong reasons as to why we need to be clear about what we mean in relevant terminology.

Another key trend across the papers is a focus on the conceptualization of the complexity of managed retreat. For example, Dedekorkut-Howes et al. [65] provide an informative review article that focuses on coastal flooding and sea-level rise adaptation practices. They undertake a cost-benefit analysis to evaluate these practices, finding that a range of adaptation options (termed structural and non-structural) are needed, and that these ought to be tailored to local needs. Mach & Siders [66] provide a powerful essay in which they conceptualize a ‘roadmap’ for managed retreat process; cognizant of human values and the necessary trade-offs that will occur as managed retreat is implemented across a range of geographies. They discuss the tensions arising out of loss, and how societies might change, retain or be transformed as a result. Siders et al. [67] also delve deeply into the many facets of managed retreat, including scale, the range of actors, the rate and scale of change, justice consideration and the implications of all these factors for both the conceptualization and practice of managed retreat. In introducing a special issue on managed retreat (the special issue itself a ‘must read’ for scholars and others interested in the field), Siders & Ajibade [68] discuss the important social and environmental justice considerations that arise in the context of managed retreat, where retreat is framed as a climate adaptation solution (see also [57,58]).

(b) . Governance, policy and institutional factors

A range of papers makes unique and insightful contributions to managed retreat literature relating to governance, policy and institutional factors, with policy and institutional planning for managed or planned retreat shown to be frequently polarizing [23]. The vast majority explore these factors at the state scale via localized/regional case study examples. This trend is reflected across a range of geographies, including the United States [57,58,69–78]; South America [79,80]; Canada [81,82]; Western European countries including France, Germany, Austria, and Sweden [83–93]; the United Kingdom and Ireland [56,94,95]; and the South Asia and Asia-Pacific region [20,21,52,82,85,96–107]. Almost all of these papers are concerned with understanding retreat from coastal or low-lying areas, with the exception of some of the USA papers where studies have explored retreat in the aftermath of significant events (for example, flood events in southern states, or the response to Hurricane Sandy [108–111], or where the focus has been on resilient urban infrastructure (e.g. [112]).

Some of the returned papers explore the legal and political tensions arising in attempting to implement managed retreat via land use planning or other legislative frameworks [21,22,29,32,113–115], with findings including challenges with community acceptance, prioritization of private property laws and rights, and political barriers to implementation via land use planning systems (see also [90]). Rocle & Salles [116] offer a unique and informative study of experimental planned retreat in which local councils/municipalities were asked to simulate retreat decision-making practices between 2012 and 2015. The results informed new legal and economic instruments, as well as provided a range of policy inputs for local planning and governance. Planned retreat, being concerned as it is with the pre-emptory process of planning for the eventual retreat from an at-risk location, will demand recognition of the cultural influences and attachments to coastlines that are going to change as climate impacts are felt. To this end, Orbach & Millar [117] posit that the cultural and policy spaces of coastlines are going to change dramatically. Using the Californian coastline as an example, they argue that coastal practitioners are required to learn new ways of coastal management to address these transformative changes, ways that are cognizant of the trade-offs and losses that will arise.

Haasnoot et al. [19] provide a compelling and detailed analysis for dynamic adaptation pathways approaches to coastal management as a response to now inevitable sea-level rise, and Haasnoot et al. [118] examine general path dependency on coastal adaptation, finding that the number (range) of adaptation options declines the more that the sea-level rises. Lawrence et al. [119] discuss planning law and governance approaches to implementing managed retreat, identifying gaps in framing and in implementing. Studies showcasing New Zealand government and governance processes are also undertaken by Christina Hanna and colleagues, and are particularly detailed and interesting. These include Hanna et al. [21], in which the authors propose a retreat governance continuum as a framework that can determine where exactly in government hierarchies' responsibility for managed retreat lies and what, if any, tools are available to implement the associated policies. Hanna et al. [20] canvass the impact of uncertainty across decision-making spheres, finding that a focus on any inherent uncertainty can paralyse managed retreat action. In this paper, they posit a refocus on forms and ways of knowing and on coordinating interactions between the governance spheres. Hino et al. [69] provide a comprehensive evaluation of managed retreat, using international cases to evaluate drivers, barriers and outcomes in terms of policy development and policy implementation. Using 27 cases of managed retreat, comprising 1.3 million people, they have developed a conceptual model that is useful for anticipating socio-political elements or factors in specific cases. Their model supports anticipatory decision-making by analysing who benefits, and who organizes, the relevant managed retreat. There is also an emerging recognition of the role that other sectors can play, in assisting governance efforts [120].

Better coordination across different levels of government is identified as a key barrier to implementing managed retreat across a variety of jurisdictions [69,92,116,119,121]. Thaler et al. [87] offer a comprehensive assessment of policy windows and opportunities to leverage those when implementing managed retreat in Austria (consider also [86]). The Thaler et al. 2020 [87] study is novel as it illustrates, over time, how relocation moved from an incremental, unplanned event to a strategic, policy supported and anticipatory occurrence. Thaler & Fuchs [88] also published an analysis that explored the connections between financial aid and recovery with pre- and post-relocation. Recognition that the motivations for needing managed retreat policy and programmes are often economic is also observed [100,110,122].

Thoughtful and repeated community engagement with respect to managed retreat policy implementation is identified as essential throughout the process of designing and implementing managed retreat, in order to ensure community acceptance of the retreat/relocation [87,123–127]. The arts/theatre plays feature as a potential communication tool [128]. Enabling community acceptance though incrementally implemented managed retreat is found to allow time for appropriate communications but also provides time for a community to adjust to the idea [129]. Cohn et al. [130] detail seven Californian case study localities, where managed retreat had attempted to be implemented but had failed. The failure was largely owing to (i) a failure of communication and appropriate community consultation and (ii) ‘baggage’ with the term managed retreat especially in terms of its relationship with property ownership. A small number of papers focus on individual adaptive capacity as a primary theme for managed retreat [131–133].

(c) . Property rights, neoliberalism and social justice considerations

Property rights and ownership are beginning to trend in systematic literature reviews concerned with coastal climate change adaptation [44] and managed retreat generally [43]. These trends are reflected across this review, with property ownership shown as an influential factor in the type of coastal adaptation, the rate of acceptance of the need to adapt, and the likelihood of property owners funding protect or accommodate adaptation options (e.g. [89,90]). Rulleau et al. surveyed residents to explore social friction with respect to managed retreat policies. Rulleau et al.'s results show overall support for managed retreat, provided that communities are involved in all parts of the process, that there is national consistency, and that the funding criteria for buyouts also be consistent and based on market price (i.e. emphasizing property rights). They explain this latter point thus:

‘…compensation for expropriated assets at their market value represents an approach based on efficiency and respect for property rights; taking into account individuals' income, and prioritizing main residences over second homes represents an approach based on need; and finally varying compensation according to whether or not individuals were aware of the risk when choosing where to reside integrates the principle of responsibility’. [90, pp. 367–368].

A study of managed retreat policy frameworks in California reveals a framework for voluntary buyout and rent-back as an option to manage competing interests in coastlines, including public access, amenity and coastal ecosystems [134]. Siders & Keenan [135] provide an innovative quantitative analysis, again United States based, aimed at measuring the relationship between types of coastal adaptation measures relative to metrics of risk. Their findings show a strong correlation between the type of adaptation option chosen and socioeconomic attributes. They ‘reveal that shoreline armouring correlates with higher home values, household incomes, and population density and low racial diversity. Property acquisitions are found to correlate with low home values, household incomes, and population density and high racial diversity’. Young [136] argues that buyouts are costly and often come too late. Young posits that governments should lease the affected land rather than buying it, arguing that this would serve the public interest and provide financial assistance in a more comprehensive and balanced way.

The financial consequences of managed retreat are heavily emphasized towards the loss of residential property. Although Phillips et al. [137] note that the economic loss spans loss of tourism as well as loss of assets and residential properties, most of the returned papers on managed retreat examine compensation through a residential property loss lens. A significant proportion of reviewed papers explored the concept of ‘buyout’, and then, mainly from the United States geography and neoliberal perspective [70,71,111,138–141] but also a New Zealand perspective [100]. The concept of buyout is where the state has provided compensation or compulsorily acquired vulnerable coastal land for the purposes of activating managed retreat. Given the emerging yet powerful property rights trends combined with neoliberal market economies and huge swathes of at risk residential and state-owned infrastructure, a key area for further research is in understanding compensation, acquisition and buyouts from a range of perspectives, disciplines, geographies and legal/policy systems [3,14,15,17,18,86,142]. Indeed, Paprocki [143] links political imaginaries with political economy in a case study of climate migration from Bangladesh to Kolkata.

Principles and frameworks deemed essential for funding planned relocation, including through aid programmes, are also emerging in the literature [144], as is recognition of the influence managed retreat policy frameworks can have in reducing or exacerbating social inequality [57,58,74,145–147]. The importance of properly assessing the equity of various frameworks for buyout implementation and outcomes is especially emphasized [57,58,71,72,146–149]. The challenge of finding suitable land to relocate to is also emerging across the literature [150]. This is leading to a rise in ‘informal settlements' [101,102,105] as people relocate themselves out of harm's way. The range of governance, policy and institutional challenges highlighted demonstrates the essential role of appropriately framed, proactive rather than reactive managed retreat policies can have in avoiding ad hoc outcomes.

Social and environmental justice barriers to managed retreat are further identified by a comprehensive comparative case study of Manila (Philippines) and Lagos (Nigeria) undertaken by Ajibade [53], including climatic uncertainties, property values and government distrust. Social and environmental justice themes feature strongly in a series of studies undertaken in the United States. For example, Schumann et al. [73] discuss resilience from an equity and Black, Indigenous and People of Colour (BIPOC) perspective, centring their analysis on the Southern States of the United States, noting that managed retreat neglects to adequately prepare for the loss of place and BIPOC attachments to specific locations (see also [151]'s analysis of the swathe of impacts funding regimes have on retreat outcomes for BIPOC communities). There are also emerging concerns that managed retreat and similar policies are being used to disconnect Indigenous communities from their land, community, cultures and traditional knowledge [75,150]. Negative social and health impacts and a need to centre human rights feature strongly in Oceania case studies [106,107,152,153], but also in the United States, where Solecki & Friedman [154] offer an analysis of the relationships between climate risks, adaptation actions and the impacts of both on well-being and mental health for coastal residents.

Nelson & Molloy [74] provide a detailed analysis of three voluntary buyout programmes in the United States and found a number of social and environmental justice factors that require consideration in managed retreat. They conclude that wealthier countries are more likely to have voluntary buyout schemes, that within such schemes, BIPOC and in their study, specifically Black people, are marginalized in such programmes for reasons of discrimination and access, and institutional and structural factors among others. They also find that the intended objectivity in the institutions and programmes distributing the aid is having a perverse effect on marginalized communities, by exacerbating injustices. Nelson and Malloy's paper demonstrates the importance of not only identifying these trends, emerging as they are across the literature, but also the value and importance of examining institutional injustices through a critical feminist lens (see further [142]). The need for (affordable) housing provision for receiving communities to adequately house people who have relocated is discussed in Li & Spidalieri [155], while Dundon & Camp [156], in their study of renters in vulnerable areas, draw attention to an often-forgotten group of affected persons (renters).

Risk perception emerged through the literature as a factor with predominately economic motivations, but overall resilience outcomes [83,84,91,110,132]. In Correll et al. [110]'s qualitative study, respondents were cognizant of economic benefits as well as avoiding flood risk in self-assessing receptiveness to managed retreat, while Seebauer & Winkler [84] analyse risk perception and its role in influencing resident decision-making in affected households, finding that, among other things, risk perception is relative to economic concerns, which in turn are framed with respect to impacts on children's futures.

(d) . Interventions that manipulate coastal environments

A range of papers discussed managed retreat in terms of either protection or accommodation manipulations of coastlines. These range from the ‘hard’ engineering structures that are argued as necessary for an orderly managed retreat [157–159], to the soft engineering options such as beach renourishment or sand dune protection [59,129,160–164], the latter of which are found to be beneficial in slowing the need for managed retreat through the rehabilitation of shorelines [130,165–167]. Hard options are found to not always be effective protections against sea-level rise over time [158], and the impacts of development on coastlines [168–171] are shown to negatively affect both adaptation options and coastal ecosystems. This has flow on effects for how hard engineering options are evaluated where ecosystem degradation is a high risk. Bonnett & Birchall [170] analyse coastal adaptation options and conclude that in some localities, a reliance on hard engineering coastal management measures is neglecting to ensure adequate consideration for more diverse forms of adaptation, including that of managed retreat. Modelling comparing beach renourishment versus managed retreat by Cutler et al. [172] offers a financial viability threshold for coastal protect and accommodate vis-à-vis managed retreat options. In addition, nature-based solutions are shown to provide economic and environmental benefits to coastal locations while also enabling a staged managed retreat [173–175].

Song et al. [176] explore coastal adaptation in Florida, USA, and find better outcomes for residents if planned retreat strategies are available sooner and offered as part of a wider range of coastal protection options (i.e. along with sand nourishment, etc). Finally, the importance and utility of land use planning policies and tools to promote and safeguard wetlands and other ecosystems is discussed in depth by Prahalad et al. [177], using Tasmania, Australia as a case study.

(i) . Where to for retreat—gaps and future research trajectories

Managed, and/or planned, retreat is a hard conversation to have, and hard policy to implement, but do it we must. This review has found a number of dominant themes arising in the literature. The dominance of governance, policy and institutional factors across all geographies is indicative of both the enormity of the retreat challenge and of the tension between a retreat that is planned before an event, and a retreat that is managed after an event. Relatedly, the points raised by papers on the differing terminology of managed retreat are an important one. First, many relevant and well-known papers published in the past 5 years were not captured by this systematic literature review. This is because authors are not describing, in the abstract or title, or tagging in their keywords, managed retreat even where those papers were concerned with managed or planned retreat. There will be specific and, in many cases, unique elements of both managed retreats and of planned retreats, each of which will require specific attention. Given the risk of fragmenting the field, the development of a typology of elements of managed retreat and of planned retreat is a clear and urgent gap to be addressed in the literature.

The field and study of managed and of planned retreat would do well to build on the recognized scholarly trajectories of managed retreat; further, to parse managed retreat and planned retreat, and to treat pre-emptory/anticipatory/strategic/etc retreat as a planned retreat with managed retreat taking place after an event. In recognizing and addressing the baggage that can arise when communicating managed or planned retreat to (a) broader public(s), the focus ought be on the style, approach and process of communicating the message rather than changing the terminology. This is not only because changing the terminology is fragmenting the academic field but also because alternating between terminology where it each means the same thing (i.e. ‘purposeful, coordinated movement of people and assets out of harm's way’, following [57,58]) runs the risk of confusing the audience or worse, having the appearance of trickery.

A reasonable proportion of the growth in the managed and planned retreat field is focused on the USA. This is perhaps not unexpected given the repeated disaster events and climate change impacts that in turn have forced retreat in some USA locations prone to flood or extremes from coastal storm events. The concern with inundation related impacts is also reflected in many papers examining South Asia and Oceania region geographies, in localities known to be low-lying and vulnerable to sea-level rise. No papers examined retreat from bushfire or wildfire locations. Given the recent occurrences of extreme and catastrophic fire events in a range of locations (e.g. Australia, United States and Europe), there is a significant gap in the managed retreat literature with respect to fire risk, as opposed to sea-level rise, erosion, or flood risk. All climate change impacts may demand some type of retreat at some point in time, across a wide range of geographies.

Property rights and enabling compensation feature strongly, as does market-based interventions and social justice impacts. Indeed, as shown in this review, a large proportion of recent papers delve into social and environmental justice challenges, making important contributions to the field especially in relation to structural racism, the challenges of a reactive market-led approach to adaptation, and in avoiding ad hoc outcomes. While Aidoo 2021 [151] does take a critical stance in their analysis, no papers explicitly adopted a critical feminist lens. This is a missed opportunity, particularly as we see more critical engagement with the role of political economy in coastal climate change adaptation scholarship and increasing relevance of property rights analysis [14,17,18,142,143]. Deconstructing neoliberalism in this way offers the field an opportunity to address head on the potential for unchecked neoliberal bias to flow into what we research, when and how. Acknowledging this has wide-ranging implications. For example, Indigenous knowledge of and connection to land and place are important perspectives often lost in a neoliberal, capital market approach to adaptation generally and retreat specifically [3,21,39,142]. Increasing the range and scope of legal and policy systems considered including that of Indigenous peoples would open the field of retreat such that these perspectives not only offer useful insights, but would open a door to thinking through the implications of alternate political economies as well as centring these voices.

The broader justice and equity themes arising from this review underscore essential developments not only for managed retreat but for the broader climate adaptation and just transitions literature and policy development/implementation writ large. At once normative and strategic, just transition belies ideas of leaving no one behind as the world and society comes to grips with wide-ranging climate impacts; and yet, with retreat, we know that some places will be lost, to great pain and suffering of many (especially Indigenous persons and groups), and that trade-offs will occur. An additional and unfortunate area of growth is the emerging evidence of increasing displacement and rise of informal settlements, occurring predominately in low-lying areas. The phenomenon of people voluntarily relocating out of harm's way or being forcibly relocated from one place to another is a trend that we can anticipate having widespread social justice implications. The environmental justice perspectives relevant to better understanding and responding to this reality cannot be understated [38,39,178,179]. Specific avenues for exploring this could (i) consider how we develop and implement new economic paradigms (i.e. new political economies, beyond neoliberalism—circa [142]), and (ii) design and implement Indigenous led community empowerment approaches (to both research and to policy in retreat specifically and climate adaptation broadly).

Finally, areas of focus including health and well-being impacts, place attachment, adaptive capacity, risk perceptions, community consultation and communication all feature in the reviewed papers (as with [44]) and can be expected to be growth areas for research and for practice. As with the opportunities and potential research trajectories highlighted above, the iterative complexity of how these features interact with each other, and over time, is what demand more explicit engagement from the research community (in addition to the areas of focus themselves—[44]). Given the substantive social justice themes arising in this review, engagement with that complexity must be aided by research that is reflexive: that is, research that properly positions who is asking the question(s), how are they asking, and for what purpose.

Data accessibility

This article has no additional data.

Authors' contributions

T.O.: conceptualization, data curation, formal analysis, methodology, writing—original draft and writing—review and editing.

Conflict of interest declaration

I declare I have no competing interests.

Funding

I received no funding for this study.