OWNERSHIP AND PARTICIPATION IN LOCAL – GLOBAL PARTNERSHIPS: THE RECOVERY OF SHELTER AND SETTLEMENTS AFTER HUMANITARIAN CRISES

: Local actors affected by humanitarian crises, including disasters and con ﬂ ict, often attract the support of global actors with engineering resources to support the recovery of their shelter and settlements. Global initiatives increasingly prioritise the agency of local actors and the concept of local participation following humanitarian crises, and a critical mass of evidence of the role of local actors in the recovery of their shelter and settlements has now emerged. This study reviews this body of knowledge and considers its implications on the global humanitarian policy framework. Data have been collected from emerging literature on the recovery of shelter and settlements and extended with sampling cases of 25 crises over the past two decades. This study ﬁ nds that for the successful recovery of shelter and settlements, the participation of affected households must prioritise their ownership of the recovery process, rather than simply their level of involvement. Furthermore, a focus on local participation appears to be most successful when tailored to the capacity households have to contribute and the shape of their plans for recovery. At the level of global humanitarian policy, a shift in priorities is required if local ownership and successful recovery is to be achieved. These priorities include more broadly assessing local capacity during the immediate aftermath of crisis, paired with systematically funding beyond the ﬁ rst 12 months after a crisis with adaptive funding instruments. The focus of this paper is the recovery of shelter and settlements, but the analysis could be used more generally in other contexts where the need for rapid global response with ﬁ nancial, technical and engineering knowledge coincides with the need to work with local social, political and technical knowledge and experiences.


INTRODUCTION
In the immediate aftermath of crises, while the majority of affected households act on their own to address their shelter and settlements needs (Parrack, Flinn & Passey, 2014;Davis & Alexander, 2016;Flinn, Schofield & Morel, 2017;Harriss, Parrack & Jordan, forthcoming), the most vulnerable groups have neither the resources nor the capability to recover and therefore often need the support of global actors.Where local actors engage global actors for support, it is widely considered that a degree of local control improves the success and effectiveness of recovery from crises (Davis, 1982;Leon et al., 2009;Karunasena & Rameezdeen, 2010;Harriss et al., forthcoming).In addition, global support that fails to align with local recovery strategies has been linked to an increase in hazard vulnerabilities (Schilderman & Lyons, 2011;Maly, 2018).
While 'local actors' can include all those living in the affected country, this paper uses it to describe communities, households and individuals who are recovering from crises.'Global actors' can include donor states, the United Nations (UN) and international nongovernmental organisations (INGOs), and this paper uses it mainly to describe INGOs.The term 'recovery' is not used in this paper to describe a phase but to describe the whole period from the immediate aftermath of crises to the point where local actors can state that they are recovered.The reason for this is because the local experience following a humanitarian crisis is a continuous process.
This study begins by considering where local actors appear in the normative framework of humanitarianism that guides the objectives and activities of global actors.This is followed by a brief history of approaches to the recovery of shelter and settlements involving partnership between local and global actors.New evidence is considered concerning the nuances of local-global partnerships for recovery, with respect to measures of success, priorities and time frames, and finally, the influence of the enabling environment provided by the global humanitarian policy framework.

RESEARCH APPROACH
Data for the study were collected through a combination of a literature review alongside discussions and interviews with humanitarian practitioners.The literature review focused on recovery of shelter and settlements following humanitarian crises in general and more specifically on the role of local actors in these scenarios, plus the policy documents of humanitarian donors.The practitioners commented on their first-hand observations in the field during a range of humanitarian crises over the past two decades.The data were analysed to identify patterns in the practice of local and global partnerships for recovery and the policy issues that influence these partnerships.
The concept of participation is set out in Arnstein and Choguill's 'Ladders of Participation' (Table 1).These ladders describe the participation of local citizens (Arnstein, 1969) and local communities (Choguill, 1996) in social programmes, and they range from positive citizen power and support, to negative manipulation and neglect.Arnstein and Choguill promoted local participation as a measure of success.
Humanitarian policies have adopted this measure, and local actors have been recognised since the early 1990s when the UN passed resolution 46/182 to reinforce the sovereignty of affected states, and the Red Cross drew up their code of conduct with a commitment to build local capacities.The UN resolution establishes the role of sovereign states in emergencies, stating that 'the responsibility first and foremost to take care of the victims of natural disasters and emergencies occurring on its territory'.1 Article 6 of the Red Cross Code of Conduct established the primacy of local actors, explicitly stating 'We shall attempt to build disaster response on local capacities'.The article goes on to recognise that local affected communities 'possess capacities as well as vulnerabilities' and that the role of humanitarian agencies is to strengthen these capacities.In addition, Article 6 highlights, as a key to building local capacity, coordination with the affected country, its humanitarian actors and the UN.Article 7 of the Red Cross Code describes the duty to local actors less explicitly, while still emphasising the capacity of the affected community, calling on humanitarian agencies to 'involve programme beneficiaries in the management of relief aid … in the design, management and implementation'.
More recently, two key humanitarian initiatives were launched, which emphasise accountability to local actors affected by crises.The Common Humanitarian Standard, launched in 2014, draws together elements of existing standards and commitments, aiming to put local communities and those affected by crisis at the centre of humanitarian action.Subsequently, the Grand Bargain, agreed in 2017, includes as one of its aid effectiveness goals to increase support to national and local responders and a further commitment to local participation.The largest humanitarian donors (Lattimer & Swithern, 2017) require their implementing partners to support the commitments of the Grand Bargain and abide by the articles of the Red Cross Code.2In summary, local actors have experienced increasing prominence in humanitarian action and with it increasing ownership and responsibility for their own recovery from crises.114 E. Babister

LOCAL-GLOBAL PARTNERSHIPS TO RECOVER SHELTER AND SETTLEMENTS
The field of shelter and settlement is highly dependent on an understanding of the local context.It calls upon a tradition of respecting the local in any global intervention.Any built structure ultimately touches the ground and, for this reason, must respect the context in which it will exist, from the topographical to the social.There follows a summary of the main approaches global actors have used to provide local actors with access to support according to their recovery strategies (Siddiqi, 2017).These approaches have their roots in the observation of households marginalised by official housing policies whose solution was to build for themselves.This 'self-help' approach was adopted by the field of developmental housing policy, promoted by architects such as Koenigserger because it was seen as cost-effective for resource providers and affordable for those with a low income (Turner, 1983).The 'self-help' approach was translated into humanitarian assistance by Davis, a pupil of Koenigserger, via guidelines for the UN's Disaster Relief Organisation (Davis, 1982).In these guidelines, he emphasised the importance of the local community and their participation in recovering their own shelter and settlements.
Based on this principle, Corsellis and Vitale (2005) developed the approach of 'transitional shelter and settlements', which can adapt to where local actors find themselves on their journey to recovery.This approach was in part a reaction to the increasing polarisation between crises-affected groups who were able to rebuild their houses and those remaining in tents, and also the common practice of considering affected communities as passive onlookers to the recovery process by distributing tents or building houses for them with little or no involvement.Transitional shelter refers to an incremental process of recovery using any of a range of temporary shelter options.A further incarnation of transitional shelter frequently used in the immediate aftermath of crises is 'Shelter Kits' that consist of a range of items to enable households to construct or repair shelters, for instance, tools or materials, based on the assumption that households also have access to the remaining resources they can use to complete a shelter (Fredriksen, 2014).The original concept was developed by Howard and Spice at Oxfam, who documented how plastic sheeting could be used to flexibly respond to crisis-specific needs (Howard & Spice, 1989).More recently, they were further developed into shelter kits by Adlam's team at the UK DFID and into a range of construction and tool kits under Saunders' team at the Red Cross.The advantage of these approaches is that they can be adapted by local actors depending on where they are on their recovery journey, avoiding 'one size fits all'.
For those who were able to reconstruct their homes, 'self-build' translated into the postcrises context as 'owner-driven reconstruction', where households were provided with funds and technical support and expected to provide construction labour themselves.Schilderman and Lyons expanded this approach into people-centred housing reconstruction (PCHR), which included a wider range of involvement for households including agreeing the strategic objectives and planning of their recovery (Schilderman & Lyons, 2011;Maly, 2018).
As global actors coalesced around the recovery of shelter and settlements, they began to borrow from other, more established humanitarian disciplines.A range of approaches emerged to guide community-based disaster risk management, including the Participatory Approach to Safer Shelter Awareness (PASSA) [International Federation of Red Cross and Red Crescent Societies (IFRC), 2011], developed under Saunders' team at the Red Cross.
Ownership in the Recovery of Shelter and Settlements following Humanitarian Crises 115 This is a variation of Participatory Hygiene and Sanitation Transformation (World Health Organization, 1997), used in water and sanitation programmes.
More recently, global actors have purposefully observed households as they recover their shelter and settlements following crises, in a similar way to Davis in the 1980s.The initiative taken by local actors has been termed 'self-recovery' (Parrack et al., 2014;Flinn et al., 2017;Maynard, Parker & Twigg, 2017;Harriss, Parrack & Jordan, 2019).One advantage of this concept is that it may apply to a range of recovery activities and at any point on the recovery journey, including construction temporary or permanent shelter.
Comparatively few studies have observed local strategies for recovery in the immediate aftermath of a crises and examined the additional support that is accessed from global actors, and fewer still focus on approaches to the recovery of shelter and settlements.Literature that considers both these aspects of recovery tends to focus on housing reconstruction, so this paper supplements this evidence with data from a range of humanitarian crises cases where other activities were undertaken.

RETHINKING MEASURES OF SUCCESS
The first challenge to the success of local-global partnerships is a meaningful interpretation of the concept of participation.While Arnstein and Choguill's ladders of participation measure success by the degree to which local actors are involved in decision making, in the post-crisis shelter and settlements context, global actors more commonly measure success by the degree to which local actors are involved in implementation.For example, owner-driven reconstruction programmes have been promoted because households can participate with hands-on control over the actual construction of building work.Despite alignment with the principle of full participation, a fully hands-on approach has not always proved appropriate for households trying to recover from crises and ultimately reduce their poverty.Two main shortcomings are discussed in the succeeding texts, namely, concentrating the resources for recovery too exclusively on small groups who have the capacity to be fully involved in their own recovery (Lyons & Shilderman, 2010;Maly, 2018) and the rejection of recovery activities that communities cannot enact themselves, including risk mitigation.

Exclusivity of Participation
The first shortcoming is the failure to include the recovery of the most vulnerable households.Those who are able to partner with global actors to implement the recovery of their shelter and settlements hands-on tend to be those with a certain level of capacity and resources, for example, they own land to accommodate a house and livelihood activities within their settlement, and they possess skills to repair their damaged assets.The most vulnerable households who have neither these capacities nor resources are therefore excluded.Leading up to the end of the 20th century, households with these capacities could be found more easily in rural locations where many crises occurred, but the 21st century has begun to see crises located more often in urban areas, where many households rent or squat3 (Earle, 2016).Urban households have been prevented from securing global support to recover their shelter and settlements because they are not owner-occupiers, and global actors are yet to develop urban approaches to humanitarian crises (Schilderman & Lyons, 2011;Maly, 2018).In these situations, the standard package of assistance offered by global actors may not fully meet their needs.(Maynard et al., 2017).Different forms of occupation call for different definitions of 'involvement' if the most marginalised households are to benefit from participation in recovery from crises.This is especially key considering the aim of both Arnstein and Choguill's ladders is social justice, because the most marginalised are those who are unlikely to be able to build their own homes.

Calibrating Participation against Capacity to Participate
In addition to different types of involvement, the level of involvement and decision making can be a burden for marginalised households.In the aftermath of humanitarian crises, households have often lost more than just infrastructure that requires an engineering solution.This can include family members, possibly the breadwinner and sometimes more than one generation of the same family. 4For these households, it is questionable whether the requirement to be involved and make decisions will have a positive and holistic impact on their recovery in the immediate and even longer term.Rather than making the assumption that participation in itself should be a measure of success, perhaps the key first decision in any local-global partnership is what level of participation local actors may actually want and be able to take on at a given point in time.

Participation versus Engineering
Using hands-on involvement as a measure of success can also work against building resilience, in the context of engineered risk mitigation.There is a balance to be struck between prioritising hands-on involvement and prioritising engineered solutions.On the one hand, global humanitarian action can be seen as a technocracy in which engineering solutions delivered through global technical assistance have gained supremacy to the exclusion of the voice of local actors.On the other hand, social justice can be seen as opening up access to the 'best' global technology in order to increase resilience and reduce poverty.At the former extreme, the participation of local actors is taken as a goal in itself and the criteria for deciding upon the nature of global support offered.Certain activities are necessarily excluded by virtue of whether local actors can implement them, not whether they will increase resilience and reduce poverty.For instance, the construction of individual houses by householders may be acceptable, but the construction of infrastructure by external contractors may not.The only acceptable activities for partnerships between local and global actors become reduced to those within the immediate sphere of the capacity of local actors to implement them.In some cases, global organisations have stopped engaging in certain resilience-building activities where expertise external to local communities is required because households cannot be fully hands-on, despite the evidence that these measures save lives.This approach has seen INGOs abandon a programme to build cyclone shelters in Madagascar because it required national contractors, and rejecting earthquake-resistant housing features in Pakistan that require engineered solutions that households could not design themselves. 5

Ownership as the Measure of Success
One possible explanation for the failure of exclusively hands-on involvement as a measure of success, drawn from the examples earlier, is that ownership of the recovery process is more important to households than the extent to which a household is physically involved with construction.Clues to this can perhaps be found in Davis' (1982, p. iii) observation that 'shelter is a "process" rather than a "product" and that "a house is merely the end product of a long chain of social, economic, technological, environmental, political and other interactions" '.Davidson et al. find that where global actors engage local actors at the later stages of the recovery process, for example, hands-on involvement in construction, this can result in more problems than local engagement in the upfront stages, such as the design of the process.This suggests that the power to make decisions can be as powerful as hands-on involvement and that local input to decisions made early on can be more beneficial because households feel ownership at a higher level.
Alternatives to using the hands-on indicator involve breaking down the process of construction to identify the different decision-making activities and the range of different types of involvement.PCHR models this approach, mapping the multiple scales of housing recovery that can be owned by households, where participation can take place: '1) policies and development at the overall disaster-area scale; 2) participation in decision-making processes at the community scale, and 3) housing design, form and construction of individual houses at the household scale' (Maly, 2018, p. 87).The self-recovery approach also prioritises household choices, observing leadership and ownership by households themselves, where they determine the type of activities they may be able to take on (Flinn et al., 2017).Community Based Disaster Risk Management in the form of PASSA allows for households to voice their recovery plans and to choose their recovery pathways and their preferred type of involvement (IFRC, 2011).These approaches may encourage global  While useful theoretical models for social justice, in the context of recovery from crises, Arnstein and Choguill's ladders could be expanded in three dimensions to include a range of participation types that encourage ownership (Table 2).For example, the top rung of both ladders can be broken down into various activities that could all lead to a level of control and ownership in different situations for different local actors.For instance, where construction is required, it can be controlled through its strategic objective, its management, its design or its implementation.Where some households may be able to manage human or financial resources, others may find this a burden but feel equally empowered solely contributing at a more strategic level.

RECALIBRATING AGAINST LOCAL PACE AND TRAJECTORY
A second set of challenges to the success of local-global partnerships is a mismatch between the recovery goals of local and global actors and at the same time the speed at which they are able to work.Recovering households have both immediate goals, such as maintaining health and re-establishing livelihoods, and longer-term goals, such as resisting hazards and reducing their poverty.Along the recovery journey, the ideal of a resilient future must be balanced against the local day-to-day goal of survival.Twigg et al. (2017) observe the shifts in these goals, using case studies in Nepal after the 2015 earthquake and the Philippines following typhoons in 2013 and 2016.Based on the voices of affected households themselves, they build a picture of recovery as a process where households' short-term priorities constantly change as households head towards their longer-term goal.They find that households do not start at the same point on the recovery journey nor do they recover at the same rate.The speed at which local actors recover can also depend on the environment, for example, whether markets are working or government policy enables recovery activities.While some communities may wait for support to arrive, especially where they have previously experienced crises, where resources are available, the majority will act alone and quickly to recover their shelter and settlements (Davis, 2016;Harriss et al., forthcoming). 6Against this fluctuating journey, the global humanitarian system still works to a structured and phased approach, with an initial urgency for lifesaving, often at the exclusion of longer-term recovery processes.The result can be that global actors find themselves catching up with decisions that local actors have already made.Those who are left behind because they lack resources and capacity may not represent a large enough group to warrant the attention of global actors.

The Tyranny of Urgency: Local Actors Move at Their Own Pace
Recent evidence demonstrates that recovery from crises usually takes several years, for example, earthquakes (Platt, 2017) and displacement by conflict (Cabot Venton & Sida, 2017).Despite this, in the initial aftermath of crises, global actors often subordinate the process of working towards longer-term recovery objectives and prioritise short-term lifesaving objectives.In general, this narrow approach reduces opportunities to respond to 6 Also discussion held at the Global Cluster Meeting, Geneva, 2018.
Ownership in the Recovery of Shelter and Settlements following Humanitarian Crises 119 the local context.Tag-Eldeen (2017, p. 409) describes this challenge, where 'the tyranny of urgency governs the conventional approaches to humanitarian assistances and limits the scope for collaboration, underestimates and underutilises local capacity and knowledge'.Furthermore, irrespective of the type of local participation or the level of control, the timing of when local actors secure global support affects the success of their recovery (Davidson et al., 2007;Harriss et al., 2019).These dimensions include the timing of initial engagement between local and global actors following a crises and the timing of local contributions in the recovery process.Even where local actors can engage global support immediately following a crisis, the timing of local contributions to the design and implementation of the process of recovery can be critical to the ease or difficulty of recovery (Davidson et al., 2007).

Prioritising Risk
Where global actors fail to catch up with local decision making in the immediate aftermath, opportunities can be lost to influence longer-term recovery.Where markets are working and households can access construction materials, they often act to meet their immediate shelter and settlement needs without adapting traditional designs to withstand hazards and reduce disaster risk.Unfortunately, local self-construction has been found in some locations to be more hazardous when left unmonitored by those with resilience expertise (Coburn & Spence 2002;Green, 2008;Parrack et al., 2014;Harriss et al., forthcoming).For example, following the Haiti Earthquake of 2010, while global actors were still focussed on lifesaving, informal markets recovered extremely quickly.Unfortunately, their supply included materials recycled from destroyed buildings, such as reinforcement bar, which were unsafe to reuse. 7Following flooding in Pakistan in 2010, households could purchase safe materials but did not apply the safer building techniques promoted by global actors and the national government following the earthquake in 2005.
The prioritisation of risk, alongside getting the timing right, is a further area where local and global approaches and values may differ and especially with respect to engineering.Despite hazards being the immediate cause of humanitarian crises, mitigating their risk is not necessarily the top priority for those trying to recover.Introducing engineering solutions such as safer building techniques can be unsuccessful if they appear to prioritise avoidance of risk over other more pressing objectives such as livelihood recovery.The prioritisation of risk mitigation may be adopted more readily if presented among a collection of broader local objectives, for instance, using approaches such as PCHR (Maly, 2018).For this approach to be successful, local actors also require sufficient resources to cover any additional costs of safer techniques to avoid competition with more immediate livelihoods strategies.There are also additional, practical reasons why building resilience to hazards remains a lower priority including the extra costs of adaptation (Tuan, Giai & Mulenga, 2014;Harriss et al., 2019), the level of certainty over government policies and the availability of labour (Maynard et al., 2017).
The broader discussion mentioned earlier, which sets up the technocratic culture of global aid organisations against local values and beliefs, is also relevant to prioritising risk.It can be a challenge, on the one hand, for global actors to respect the autonomy and agency of local actors and on the other effectively promote a more resilient future desired on their behalf.Johnson, Wahl and Thomalla (2016, pp. 2, 4) highlight that with the proliferation of global actors involved in humanitarian crises, 'the convergence of multiple cultural responses to risk formed in distinctly different contexts can lead to misunderstanding, compromising the effectiveness of DRR efforts' and call for global actors to 'understand various perspectives, values, and motivations that shape decisionmaking processes in times of uncertainty'.Consideration of broader local beliefs, values and objectives may contribute to a better understanding of where local actors are on their recovery journey and to what extent they are willing or able to partner with global actors to be involved in tasks to recover their shelter and settlements.
In summary of the above, for local actors to engage meaningfully with global actors, local actors need time and voice to more clearly communicate their recovery plans, and global actors need an appropriate conduit to provide resources.Firstly, the timing and type of engagement appears to be key.Early engagement is beneficial alongside a partnership that supports local decision making at a strategic level to ensure ownership of recovery.Secondly, understanding and respect for local beliefs and values, including immediate and long-term priorities related to perceptions of risk, appears to contribute to more successful recovery in the longer term.

Can We Prepare the Longer Path?
Partnerships between local and global actors supporting recovery of shelter and settlements following crises necessarily exist within the international humanitarian system.For local actors to move at their own pace and retain ownership over their recovery, the international humanitarian system needs to be an adaptive enabling environment and one that makes room for the consideration of longer-term recovery plans in the immediate wake of a crises.A brief review of the current key elements of the humanitarian enabling environment provides clues to why local-global partnerships have so far been challenged to claim this space.Elements of this enabling environment include the global coordination architecture in the form of the Cluster System, the data collection process in the form of needs assessments and a funding system that reinforces these approaches.Recent initiatives such as the Common Humanitarian Standard and the Grand Bargain provide hope as they attempt to promote a certain level of local participation, but more systemic changes are required to how global actors manage humanitarian information and resources if local ownership is to be achieved.

Early Recovery
In the main, the Cluster System is built around saving lives, and this is demonstrated in the mandate of the co-leads of the Global Shelter Cluster, IFRC and United Nations High Commissioner for Refugees, whose mandates are to lead the cluster only until 6 months after a crisis.While the Cluster System is largely built to enable a short-term response, the idea that the design of recovery processes should be supported immediately following a crisis existed formally, but briefly, in the form of the Early Recovery Cluster, from 2005 to 2018.The Inter-Agency Standing Committee Principals conceptualised early recovery 'a multidimensional process of recovery that begins in the early days of a humanitarian Ownership in the Recovery of Shelter and Settlements following Humanitarian Crises 121 response'8 and requested that the majority of clusters integrate early recovery into all the different phases of the Humanitarian Programme Cycle 'as the foundation for strengthening resilience in a crisis or post-crisis context'. 9The Early Recovery Cluster operated mainly as a network led by the United Nations Development Programme that reached out to sectoral clusters as a resource for the early consideration of recovery.The formalisation of early recovery signalled recognition by the humanitarian system of the value of starting to plan recovery early and aligned global support more closely with the pattern recovery taken by local actors.To a certain extent, the existence of the early recovery cluster has validated both funding and engagement in longer-term approaches to shelter and settlements recovery such as the consideration of housing, land and property rights, for illustrated by the establishment of the Shelter Cluster's Housing, Land and Property Working Group following the Haiti Earthquake of 2010.
Latterly, early recovery has been hindered by conceptual confusion.It is unclear whether the expectation is to ensure humanitarian actors take a more developmental approach or that rehabilitation should just take place alongside immediate lifesaving.With this failure to be precise, early recovery has therefore been considered as an administrative creation of the aid bureaucracy, rather than a new analytical tool (Mosel & Levine, 2014;Murray, Pedersen & Soenke, 2018).This has ultimately led to a recommendation to downgrade early recovery from a cluster to lighter mechanism in the form of an advisory network.Perhaps any concept placed in the silo of a cluster in a coordination system that has been criticised for a lack of local access will be insufficient to create space for local recovery plans at the level of humanitarian architecture.It remains to be seen whether the Grand Bargain commitments around localisation and participation create more sustainable windows of opportunity.

Needs Assessment versus Capacity Assessment.
The second element of the humanitarian enabling environment that can work against localglobal partnerships is the process of needs assessment.Needs assessments can be powerful because they are generally used the fundamental justification for global humanitarian response strategies, and these in turn are used to justify funding appeals.Engineering culture in particular tends to focus on what is broken, for example, damaged buildings, and what is needed to repair the problem, for example, material resources.While these data are important, conceptually, it gives a partial picture because it narrowly defines the local context as broken and the path to recovery as defined by resources to fix the problem.Data from needs assessments speak of what local actors need rather than giving voice to their recovery plans or auditing their capacities.On occasion, this can lead to missed opportunities and wasted resources.After the 2004 South Asian Tsunami in Sri Lanka, those living near the beaches were assumed to be fishermen, so some were given boats, when many of them were actually construction workers. 10More recently, needs assessments have become more sophisticated to include data about the behaviour of local actors, for example, the Multi-Sector Initial Rapid Assessment (Inter-Agency Standing Committee, 2015) asks 'How do different groups cope with the emergency situation?',but even this does not go as far as asking about how different groups intend to recover or what they can contribute.
A couple of recent advances in assessment where a broader data sets are gathered include market assessment and cash programming.Markets assessment considered local markets for what and how they can contribute to humanitarian response.The markets approach was born out of the food security sector (Byrne et al., 2013) and therefore considers commodities.Shelter and settlements actors now collect markets data, for example, looking at bamboo in the Royhinga Crisis in Bangladesh (Martin, 2017), but more recently, they have begun to consider its application to assess sources of labour, which could include labour capacities of affected households themselves.11At the same time, the rise of cash programming (Bailey & Harvey, 2015) has necessitated an approach to assessment that must consider household capacities with respect to financial management, and this approach could influence humanitarian data collection to include the capacities local actors can contribute.Where households are asked about their recovery plans and how they intend to manage and resource them, the answers would build not only a picture of local processes but also the time frame over which they expect to achieve their recovery.These types of questions are also likely to reveal capacities at household level that may be used in the recovery process at household, community and market level.

Funding
The two elements above, namely, coordination architecture and data collection, are both fuelled by and therefore influenced by funding.The flow of the majority of funds is often aligned according to an international priorities rather than local priorities because resources generally flow from the international humanitarian system to the country that has requested international support (Lattimer & Swithern, 2017).Local-global partnerships can therefore be frustrated by the supply of resources.Challenges can include the scope of what can be funded, the timing of funding, the time frame of funding and the quantity of funding.
The current funding system reinforces the short-term humanitarian response paradigm and a short window for needs assessment.Because the global focus in the aftermath of humanitarian crises is on lifesaving, so this leaves little room for funding local actors to engage with global actors on the broader recovery process.While the humanitarian policies of the larger donors, for example, government donors, value activities beyond lifesaving such as resilience, this type of activity is likely to be short lived given that the time frame for initial humanitarian grants is often less than a year (European Civil Protection and Humanitarian Aid Operations, 2013; UK Department for International Development, 2018; Office of US Foreign Disaster Assistance, 2018).Opportunities to build long-term local-global partnerships or to use approaches such as PASSA are challenged by this short term culture at system level.
Short spending windows of humanitarian funding appear to be intractable obstacles to recovery.Even where a large quantity of funds is experienced in the immediate aftermath of a crisis, the rigidity of donor spending restrictions prevents the funds being spread along a longer time frame.For example, some INGOs underspent their humanitarian grants in the first 6 months following Typhoon Yolanda, in the Philippines, but the donors would not let surpluses be spent on longer-term recovery programming (Stoddard, Harmer & Hughes, 2015).Following the 2015 earthquake in Nepal, INGOs felt pressure to spend funds in the first year, despite evidence that the cost of recovery would rise over the coming years. 12here households do have long-term recovery goals, but where funding is limited, they often experience a shortfall between the assistance provided and their goals for recovery (Twigg et al., 2017).A further challenge, caused in part by short funding time frames, is the difficulty in retaining external interest and support for the recovery of shelter and settlements after emergency needs have been met (Davidson et al., 2007).
Local recovery plans require funding that can be released quickly and that can accommodate adaptive planning processes and longer time frames.Two funding innovations show promise in this respect, although neither are yet systematically used in the immediate aftermath of crises.The first innovation is multi-year humanitarian funding (Cabot Venton & Sida, 2017;UN Food and Agriculture Organization, 2017) where grants are made available for several years.This has been established by the UK Department for International Development in protracted crises, including in Ethiopia, Sudan, the Democratic Republic of Congo and Pakistan, but is yet to be applied to rapid-onset contexts.The second innovation is the Crises Modifier (Feinstein International Center, 2015;Fabre, 2017;Rohwerder, 2017), which is a provision in a grant agreement that allows the organisation to move funds from development activities to crisis response and may also allow the donor to provide additional funds for crisis response, without modifying the grant agreement.Originally designed to support slow onset food insecurity, these have been trialled in rapid-onset contexts by United States Agency for International Development in Africa and Haiti.

CONCLUSION
Several policy issues arise from this review of building local and global partnerships for the recovery of shelter and settlement after humanitarian crisis.The policy issues are specific to shelter and settlement but could be generalised to other situations where the opportunity to access rapid global technical engineering needs is open to local actors.These areas could include the recovery of water and sanitation infrastructure, transport systems or medical facilities.
This study finds that for the successful recovery of shelter and settlements, participation must mean local ownership.While the global humanitarian policy framework strives to respect the voices of local actors, informed by the concept of participation, participation as involvement alone appears to be insufficient without early engagement between local and global actors and a shared understanding of immediate and longer-term recovery priorities.
Furthermore, a focus on local participation appears to be most successful when tailored to the capacity households have to contribute and the shape of their plans for recovery.Local actors will be able to use global engineering resources to a greater advantage of global actors slot into these plans in terms of speed, timing and design.Participation, and particularly hands-on involvement, as a lone measure of success can result in further marginalisation of vulnerable groups who may not have the capacity to participate.
Proactively creating the space for listening to local recovery plans and building local ownership must be prioritised within the global humanitarian policy framework if vulnerable groups are to successfully recovery their shelter and settlements after humanitarian crises.This means going beyond needs assessment to more broadly assess local capacity and systematically funding beyond the first 12 months so that recovery plans can be made and shared with a degree of confidence.

5
Strategic discussions with INGO Country Directors in Pakistan 2010 and in Madagascar 2008.

Table 1 .
Ladders of participation

Table 2 .
Types of participation and capacity to participate Examples of types of participation compared with the capacity required to participate.118 E. Babister actors to recognise the range of capacities households possess and to propose a more appropriate range of different types of support.