PCR POSITION PAPER Putting People at the Centre of Reconstruction Michal Lyons and Theo Schilderman The overriding aim of reconstruction programmes should be to make people more resilient to future risks and change. That requires both making their buildings more resistant and safer to live or work in, and helping people themselves become more capable of adapting to risk and change. Where it comes to housing, building back better is narrowly interpreted by many agencies as reconstructing houses that are more resistant to disasters than previous types. That concern for quality is a key reason for them to prefer reconstruction designed by architects and engineers, and built by contractors. Such an approach is not only expensive, but does little to make people become more resilient. Moreover, if standards are set too high, they may be unaffordable for people to maintain in future. Instead, it is entirely possible to support and empower people to take charge of their own reconstruction processes and realise adequate quality. Assessments of how housing behaved in a disaster often point at vernacular technologies with a good record, such as dhajji dewari in Kashmir, 2005. Where they show any weaknesses, these can usually be easily overcome. Adopting such improved local technologies requires less support for reconstruction, and is usually cheaper and quicker. Building better houses, however, must go hand in hand with rebuilding people’s livelihoods and local markets as well as their social networks and must allow survivors as large a role as possible in decision making and management of resources. This strengthens their capabilities to cope and adapt. Above all, the process of reconstruction matters as much as its end products: putting people at the centre will empower them and strengthen their capabilities and resilience. Until recently, such methods, generally known as Owner-Driven Reconstruction (ODR) have been exclusionary, largely focusing on people with existing title to land and housing. This paper advocates that an inclusive and participatory approach, PeopleCentred Reconstruction (PCR) should be at the heart of all approaches to housing and livelihoods after disasters. Currently dominant approaches to reconstruction, which are centrally planned and executed, frequently fail to reach all who matter and often do not do enough to tackle the underlying causes of poor people’s vulnerability to disasters. Recent experience calls for new thinking! The Indian Ocean tsunami of 2004 was a disaster on an unprecedented scale and elicited local and international responses that went well beyond the experience of previous disasters. It required responses on a very large scale and brought new actors onto the relief and reconstruction scene. This provided a test-bed for new and established approaches to relief and reconstruction. The response to the huge Kashmir earthquake of 2005 added to that experience. The scale and frequency of disasters have increased in the past and are likely to increase further, as a result of e.g. climate change, urbanisation and persistent poverty. The increase in hydrometereological disasters is particularly significant. It is essential to improve the capabilities of communities and institutions alike to both mitigate risks and build back better. Natural disasters are becoming more frequent and are affecting growing numbers of people 19701979 Hydrometereological Geological Biological Total 776 124 64 964 19801989 1498 232 170 1900 19901999 2034 325 361 2720 20002005 2135 233 420 2788 Source: UN-ISDR, after EMDAT http://www.unisdr.org/disaster-statistics/ Table 1. Distribution of natural disasters by origin (1970-2005, by decades) Over the same period, the reported number of people affected by natural disasters rose fivefold from an average 45 million per year around 1970, to 230 million around 2005, according to the OFDA/CRED International Disasters Data Base, EM-DAT. (http://www.emdat.be/natural-disasters-trends) People-centred approaches can help achieve this through both product and process. The links between development, vulnerability and disasters were first clearly identified by Wijkman and Timberlake in 1984. Disasters of similar magnitudes have much greater impact on poor countries than rich ones and, within countries, the poor are generally more affected than the rich. The poor in developing countries are particularly at risk of disasters, because they lack the resources to build better, do not always have the right skills, knowledge or information, and often have to make do with dangerous sites to build on, e.g. steep slopes at risk of landslides, or alluvial plains at risk of flooding or liquefaction. And amongst the poor, women are more at risk than men, because they often have less of a voice, own less assets, have less physical strengths, and are sometimes hampered by their clothes. Understanding how and why people built as they did before a disaster can highlight underlying vulnerabilities and capabilities, as well as strengths and weaknesses in construction. Some of the most successful reconstruction projects and programmes adopted vernacular skills and technologies with a good disaster record. The crucial role played by local actors in disaster mitigation and reconstruction has been recognised since the 1980s by agencies including Practical Action (then ITDG), as described by e.g. Maskrey (1989), using the example of Peru. This echoed findings in the informal housing sector there from a decade or so before (Turner (1976). Both show that the people’s participation in construction or reconstruction is important; it empowers them, builds their capabilities and makes them more resilient. Notwithstanding all the losses and distress they cause, disasters generate opportunities as well, particularly if they are large scale and lead to a significant humanitarian response. The ensuing influx of external finance generates ample scope to address some of the weaknesses and resuscitate local markets and livelihoods. Unfortunately, such opportunities are often wasted in practice. Agencies will need to reflect seriously on the above lessons, before deciding how to rebuild in Haiti after the 2010 earthquake. We now turn to examine these issues in more detail. The poor suffer more in disasters and have a harder time recovering from them When an earthquake measuring 6.9 on the Richter scale struck San Francisco on October 17th, 1989, it killed 62 people, affected 3,757 and caused 5.6 billion US dollars of damage. A few years later, on September 30th, 1993, a somewhat weaker earthquake, measuring 6.4, hit Maharashtra in India, killing 9,748, affecting over 30,000 and causing 280 million of US dollars of damage, including 171,000 houses. On January 12th, 2010, an earthquake measuring 7.0 on the Richter scale struck Haiti, killing an estimated 222,570, and affecting 3.7 million people, many of those in the shanty towns of Port-au-Prince. Less than two months later, on February 28th, one of the strongest earthquakes ever, measuring 8.8., hit Concepción in Chile. Data are still sketchy, but most victims seem to have been caused by an ensuing tsunami; it is unlikely that the quake itself killed more than 1,000 people. All these earthquakes hit very populated areas; the huge differences in their impact are due to variations in poverty and vulnerability of the people living in those areas. The differentiated impact of disasters on men and women is primarily caused by existing gender inequalities. Statistics show that natural disasters and their impacts kill on average more women than men, and kill women at an earlier age than men. Whilst women survive as well as men in societies with equal rights, where these do not exist, they can suffer badly; e.g. of the 140,000 people that died from cyclones in Bangladesh in 1991, 90% were women. (Aguilar, 2009) More resistant housing can often be developed from local technologies Figure 1. Woman in Chincha, Peru, rebuilding her house in improved quincha after a recent earthquake © Practical Action Latin America 2 The poor are more vulnerable to disasters... The bulk of low-income housing in developing countries is built by the poor themselves. The constraints they face in doing so are largely responsible for their vulnerability to disasters and also for their vulnerability in the reconstruction process. First, the poor often build on hazardous sites, exposing themselves to flooding, landslides, settlement and liquefaction, as well as disease vectors. Such hazardous areas are often worse hit by disasters. Second, livelihoods of the poor, whether rural or urban, are often more vulnerable to disasters and, more importantly, the poor have fewer resources to recover their livelihoods. Thus, poverty compounds the effects of disaster. The poor often build their houses on hazardous sites ... but the poor are also more vulnerable in the reconstruction process Although reconstruction creates an influx of resources for development, this may not work equally well for everyone. Successful reconstruction calls for inclusive and integrated thinking. Vulnerable livelihoods are often a barrier to individuals and society recovering from a disaster. Looked at another way, housing reconstruction generally takes place before livelihoods reconstruction and often ignores its needs. If reconstruction doesn’t do enough to rebuild livelihoods, the poor may well end up worse off than before. Support in the reconstruction of housing stock from the state or international agencies is often limited to those who can prove title to land or buildings. This excludes or discriminates against tenants and squatters – groups which form a high percentage of the poor in most developing countries. When money starts coming in, it is frequently captured by an elite, people who have the influence, connections and knowledge to deal with the bureaucracies in place. Where immediately after the disaster, poor and rich were working together towards recovery, as soon as serious resources come in, this may put them in different camps, and the poor may lose their position and influence in the group. Figure 2. These houses built on unstable terraces on steep slopes in peri-urban Lima will be very vulnerable to earthquakes © Practical Action, Colin Palmer The lack of secure tenure can exclude people Post-tsunami reconstruction in Sri Lanka had two components: an Owner-Driven Programme (ODP), allowing people to rebuild on their own plots, outside a buffer zone, and a Donor Assisted Programme (DAP) for people who were living within the buffer zone and had to be relocated. The ODP targeted owners of land and property only. This created problems for tenants, squatters on private land and for extended families where titles perhaps had not been subdivided. At the end of 2008, for instance, there were still over 500 of such families living in transitional camps in Colombo district alone, faced with the risk of eviction. Whilst the government reconstruction agency, RADA, recognised this as a challenge and did propose a third phase of housing assistance to deal with it, this never happened due to the large caseload and restrictions in funding. Within the DAP, however, a different approach was taken. Persons living within the buffer zone before the tsunami were assisted on the basis of humanitarian needs alone. Thus, squatters on government reservations fared better, especially in Colombo, Batticaloa and Trincomalee. The authorities provided them either with a plot in a relocation site, or with cash to buy a plot elsewhere, after which they were entitled to support under the ODP. (Hidellage and Usoof, 2010) Fifty years of approaches to housing It is important to develop a good understanding of the housing sector in a disaster location, when preparing for reconstruction. This is in the first place because the context (including policies, strategies, rules and regulations, culture, skills, access to materials, land and finance etc.) that defines housing will also apply to reconstruction. Secondly, that same context is an important factor in how houses behave when a hazard strikes, and why that hazard may become a disaster. And thirdly, there is lengthy experience with residents playing central roles in realising 3 Well-built rural housing can often withstand disasters Mama Susan builds a house and a livelihood in town Mama Susan Taplokoi Maina lives in one of the spontaneous settlements that have sprung up around Nakuru in Kenya. Practical Action first came across her in 1997, when she was about 60 years old. She had lost her husband, and of the six children she had with him, four died young. She had managed to buy a plot with her savings and the help of others, and was living in a single-roomed house made of timber off-cuts and galvanised iron sheets. Her main income came from selling second-hand clothes and vegetables in the market. As she was getting older, she did not want to end up being a burden on her children. She needed a more secure income that could help her to improve her house. Using some of her income from sales, she extended the house to add a single room that she rented out. The rental income helped her to build another room, and another. The more rooms she built, the more her income increased. By 2003, she was letting 12 rooms, at about $7 per month. From this, she managed to raise the capital to invest in materials to build a better house for herself. For many years, Mama Susan has been a member of a self-help group, initiated to improve the neighbourhood. They also started to discuss how their houses could be improved. They finally opted to build their walls with stabilised soil blocks, and Practical Action worked with the group to build their capacity to do so. Mama Susan now lives is a much better and larger house on her own plot. Figure 3. A two storey timber-framed house such as this in San José de Lourdes, near the source of the Amazon in Peru, may resist earthquakes and provide safety during flooding © Practical Action Latin America regular housing, which reconstruction processes can build upon. In rural areas, poor people have always built their own houses, and rural housing often has a rich tradition. Housing designs and technologies have been passed on and improved upon by generations of rural builders. They often include measures to mitigate the impact of disasters, born from experience, and these must inform reconstruction. Many rural communities also have builders specialised in specific components of construction. The quality of rural housing varies with local skills and materials. Durability may be a problem, since many natural materials are easily affected by humidity and insects. Wellbuilt and maintained rural housing has proven to resist low- to medium-intensity disasters but most remains vulnerable to major disasters. Rural housing and livelihoods are often closely linked, with houses being designed to e.g. store agricultural produce, keep livestock, or accommodate agro-processing. The production of housing in urban areas is more complex. It is less easy for the urban poor to build housing themselves, because it is Figure 4. Mama Susan’s house in Nakuru © Practical Action/Liz Fredericks hard to access land, natural materials are less available, and construction is more monetised. Some therefore revert to renting and others become squatters. High rates of urbanisation have contributed to huge housing backlogs over the past decades. Governments have attempted for fifty years to tackle these backlogs by building houses for people. However standards were far too high and the urban poor could not afford them, unless they were subsidised; production costs were high and output was a fraction of the need. 4 When the limitations of public sector housing supply became clear, in the 1970s, international agencies such as the World Bank promoted the provision of serviced sites and sometimes a core house, to be completed by the owners in time. However, the sites and services programmes suffered from similar limitations and failed to reach large numbers of the urban poor. The approach was abandoned in the late 1980’s, to be replaced by approaches which see authorities as enablers of housing processes in the private sector. Urban housing processes thus changed from being ‘supply-driven’ to being ‘supportdriven’, a clear sign that authorities and agencies had begun to value people-centred housing processes. Of course, these processes have always been there, often with the urban poor in the driving seat, building the informal settlements that started to sprawl in and around towns and cities. Popular urban housing is often built over many years, in an incremental process of extensions and improvements as resources become available, providing owned and rented homes. The case of Mama Susan, is an interesting one, in that it explains how those housing processes may happen, and some people end up as landlords. For Mama Susan, her housing became an important part of her livelihood strategy; many other people will have home-based enterprises; and for others again the location of their house is important for being close to livelihood opportunities. Official strategies to improve the living conditions in informal settlements have tended to focus on upgrading the services, leaving housing to the owners. Low-income urban housing is often built without regard to standards and on dangerous sites. When disasters hit towns and cities, it tends to be more heavily affected. DDR: Similar houses for different families, located away from livelihoods Figure 5. Post-tsunami reconstruction scheme near Hambantota, Sri Lanka © Practical Action, Lucy Stevens Slowly changing approaches to reconstruction… In 1970, major earthquakes struck Peru and Turkey, causing much damage and many casualties. In both cases, the government initiated large reconstruction programmes, often involving relocation, and received assistance form external humanitarian agencies on an unprecedented scale. The approaches followed by governments and agencies alike were to build houses for people rather than with them. And they had important flaws: many of the houses built remained unoccupied, and affected people reverted to their old ways of building, remaining vulnerable to future risks (see, e.g. Aysan and Oliver (1987) and Blaikie (1994). This approach to reconstruction, which had agencies in the driving seat, is often termed Donor-Driven Reconstruction (DDR). It has been much studied since 1970, and many of those studies have concluded that DDR has important drawbacks. The most important ones are listed in the box overleaf. The overall conclusion of all these is that DDR should not be recommended, except for cases where very little local building capacity remains. Tragically, forty years on, governments and agencies sometimes still adopt centralist approaches, and houses remain unoccupied even now. Disasters do put a lot of pressure on decision makers. They lead to thousands of people living in makeshift accommodation or with host families, and nobody likes that to last for too long. If nothing happens quickly, the media will put on additional pressure. So now, as in 1970, decisions are frequently made in a hurry, without much investigation or analysis or without involving all those that matter. It appears to remain difficult for some decision makers to shift from a supply-driven ‘relief mode’ to a ‘reconstruction mode’ that ought to be much more support-driven and people-centred. A move in the right direction has been the emergence at scale of Owner-Driven Reconstruction (ODR), about a decade ago in Asia. The approach itself was not new; it had been supported largely by NGOs on smaller scales for several decades, especially in Latin America. What perhaps influenced the greater interest and scaling up are the changes in housing policies and strategies, from supplydriven to support-driven, over the years. Thus, more agencies recognised the major role played by home owners in the production of houses under normal circumstances, and queried why reconstruction after disasters should happen in such a different way. They therefore gave a much more prominent role to property owners in 5
Practical Action,