PCR TOOL 3
Learning from Disasters
In order to be able to build back better, we need to
understand what caused the hazard that occurred
to become so disastrous for the people it affected.
This involves not only studying why houses were
vulnerable to collapse, but also the underlying
causes for that; these lie in the vulnerabilities of
people themselves. What happened in the Alto
Mayo earthquake, described in the box, right,
explains why. Although a study of housing revealed
that some building technologies were more
resistant than others and that certain residents
and local builders had the capabilities to construct
well, within those technology categories there were
variations too. These often resulted from people,
who were getting poorer, being no longer able to
afford to build or maintain their houses well.
Vulnerability is now receiving more attention,
not just in the context of disaster reduction,
but also with respect to issues such as drought,
food security and increasingly climate change.
It is not enough, though, to only know people’s
vulnerabilities, or weaknesses. We also need
to explore their coping capabilities, because
these are the strengths on which to base better
reconstruction. In the case of the Alto Mayo,
one coping capability was the local knowledge of
resistant building technologies.
Vulnerabilities and assets – which include
capabilities – are key components of sustainable
livelihoods analysis, which goes back to the
thinking of Robert Chambers in the 1980s. The
livelihoods approach puts people at the centre of
development. Livelihoods analysis is helping us to
understand that poverty is multi-dimensional and
that disasters are not the only risk poor people are
facing. For some poor people, day-to-day survival
may be a greater concern than the distant threat
of a disaster. People do not willingly run the risk of
death or asset losses, but short-term pressures such
as the need to make a living or to feed a family
may force them to accept the more remote risk of
disasters. For example, a study of the Karakoram
region of Northern Pakistan from the 1980s found
houses to be dangerously located on slopes. The
owners were aware of the risks these locations
posed, but opted to build there rather than using
the little arable land they had for housing.
The Alto Mayo earthquake of 1990
When a moderate earthquake struck the Alto Mayo of
Peru in 1990, the region was in economic decline.
Its main product was rice, but the government had
disbanded the agency buying rice from farmers and
failed to properly maintain the one major road that
linked the region to the markets of the main cities
on the coast. Many incomes therefore declined;
this reduced people’s capabilities of building and
maintaining their houses well, and this proved to
be a major factor in the damage and casualties the
earthquake caused. The region’s inhabitants had
become more vulnerable because their livelihoods
had been negatively affected by external events, in
this case a government failing to do its duty to them.
What is more, when aid started to flow into the region
in the aftermath of the disaster, it included a lot of
imported rice, at a time when local stores were full
to the brim of rice farmers were unable to sell. This
further worsened their potential for recovery, as it now
became nearly impossible to sell rice locally. However,
observations of the impact of the disaster also showed
that not everybody was equally affected. Houses built
with heavy rammed earth (tapial) or adobe walls –
built by people who had migrated into the region from
Cajamarca - had generally performed badly, but those
with much lighter mud-and-pole (quincha) walls quite
often stood up. The latter technology then, with some
improvements, became quite popular in reconstruction
supported by Practical Action.
Earlier thinking on reconstruction did not
pay much attention to people’s livelihoods and
vulnerabilities. It tended to concentrate on
technical issues, e.g. the weaknesses of housing,
and how these could be overcome, often by
construction experts rather than the people
themselves. Affected people were frequently forced
to relocate away from sites considered dangerous;
some chose not to occupy the alternative houses
offered, whilst others moved out to return to places
where they could resume their livelihoods and
social networks. In other cases, new houses were
built using technologies so alien or expensive, that
inhabitants were unable to maintain or replicate
them and ultimately reverted to the old ways
of building, reinforcing their vulnerability. This
thinking is now changing, but only slowly.
It is important to understand what people’s
vulnerabilities were before a disaster struck, but to
photo © Chris Martin
Reconstruction with improved quincha after the
1990 Alto Mayo earthquake in Peru
also be aware that the disaster may have increased
them. Reconstruction will have to help tackle
vulnerabilities, and the influx of external resources
often offers a unique opportunity to do so. Building
back better should not only apply to housing, but
equally to rebuilding the livelihoods of people
affected by a disaster, and of local markets. Ideally,
that should happen through an integrated approach
or in close co-operation between agencies involved.
PCR Tool 6: Integrating Livelihoods, explains this in
How can learning from disasters help
to build back better?
The early research into how traditional housing
behaved in disasters, from the 1960s until the
1990s, was generally undertaken by technical
professionals. They would make observations
in the field after disasters occurred, perhaps do
further laboratory tests and calculations, and draw
conclusions with respect to particular construction
weaknesses. These results were communicated
largely with other professionals and sometimes to
NGOs and authorities, often in ways that are alien
to poor people. Therefore, little of this information
actually reached those affected by the disasters,
or others potentially at risk. People-centred
reconstruction (PCR) offers more opportunities
for learning than previously dominant research
and reconstruction approaches. Involving affected
people in reconstruction processes from a very early
stage, that is from the moment when damage and
needs are assessed, can lead to clear changes in
how those people perceive and understand disaster
risks, what they can do by themselves to reduce
these, and where they need external support.
In many rural areas of developing countries
there is a tradition of vernacular construction;
evolved over many generations in response to local
customs and culture, climate and availability of
low-cost local materials. This often incorporates
measures to mitigate local hazards, including for
example: special types of foundations to prevent
houses sliding down steep slopes; building on
raised plinths or on poles fixed deeply into the
ground where flooding sometimes occurs; and
bracing frames and ties to resist high winds. In
the historic centre of Lima, Peru, for instance, a
number of vernacular houses built with quincha
survived several strong earthquakes, including
those of 1746 and 1940, and remained standing
where many modern masonry buildings collapsed.
Where people abandon their vernacular
ways of building, for example, when they move
from a rural to an urban area or for reasons of
status, they are likely to also abandon some of
the traditional ways of mitigating disasters (for
instance, reinforcing roofs prior to the hurricane
season, or repairing houses after rains). Building
in rigid masonry, as in the above case of Lima, is
not inherently unsafe, but does require different
measures to make it more earthquake resistant. A
major change in construction technology usually
requires a change in the way a building is protected
from disasters. The knowledge for this does not
come automatically, nor is it easily acquired
by simple observation. There is therefore a risk
that many such buildings are poorly constructed
and unsafe. This can happen, for instance, in
urban and peri-urban areas, where people choose
to construct reinforced concrete buildings but
lack the knowledge and resources to adequately
reinforce main structural elements such as beams
or columns. Another risk occurs where people cut
corners in construction. They may, for instance,
decide to do all the building work themselves,
without hiring a skilled builder to help with the
more complex parts. Alternatively, they may try
to save on materials, for example, by using less
cement in concrete than is commonly required,
or making foundations much more shallow than
they should be. In high density suburbs, people
may decide to add floors to a building which is not
structurally capable of carrying such extra weight.
All of these factors can increase the risk of damage
to, or collapse of, houses in a disaster scenario.
The participatory process of learning from
disasters can help all involved in reconstruction
(communities, authorities, humanitarian agencies)
to better understand the strengths and weaknesses
of local construction methods, the underlying
reasons and vulnerabilities, and the capabilities of
residents and local builders.
The importance of learning in reconstruction
had been recognised for a long time. Yames
Y.C. Yen for example, founder of the Rural
Reconstruction Movement in China in the 1920s
expressed the ideals of people’s participation in
development and improvement, highlighting in
particular, the importance of learning in the process
(see box to the right).
In post-disaster reconstruction, many of the
most recent successful examples illustrating Yen’s
principle of Starting with what people know, have
begun with architects and engineers finding out
how people are building already, and, in particular,
what they are doing already to reduce disaster risks.
They then work to produce improved designs that
incorporate much of the traditional elements. In
such cases, the architects and engineers must not
assume that they know better than the traditional
builders. They must work together in partnership,
perhaps modifying the design several times before
arriving at a model all stakeholders are happy
with. Some recent reconstruction projects and
programmes, e.g. the ERRA programme in Azad
Jammu and Kashmir and North West Frontier
Provinces of Pakistan (see case 1 in the section
Applications below) have followed this approach.
What do we need to learn?
In order to build back better, stakeholders in
reconstruction jointly need to find answers to four
sets of key questions:
1. Why were people vulnerable to the hazard that
occurred? Did vulnerabilities differ amongst
various categories of people (e.g. men/women;
owners/tenants; land owners/landless; able/
disabled)? Has the disaster further aggravated
the pre-disaster vulnerabilities? Who are the
individuals or categories that are particularly
at risk and will need special attention in
reconstruction? These questions can be
answered through vulnerability assessments.
2. What made people’s housing vulnerable to the
hazard? What were the predominant building
technologies and what were their relevant
strengths and weaknesses? What factors
affected disaster resistance within single
technologies? These questions can be answered.
through damage assessments.
3. What is the likelihood of disasters happening in
this particular location? Does it have particular
geographic features that make it vulnerable?
James Y.C. Yen on Learning
‘Go to the people
Live among the people
Learn from the people
Plan with the people
Work with the people
Start with what the people know
Build on what the people have
Teach by showing; learn by doing
Not a showcase but a pattern
Not odds and ends but a system
Not piecemeal but an integrated approach
Not to confirm but to transform
Not relief but release.’
Are there any other risks besides those directly
related to the disaster? These questions can be
answered through disaster risk assessments.
4. What are the local capabilities, amongst
residents as well as builders, to build in
disaster-resistant ways? Are the required
resources for reconstruction (manpower with
the right skills and materials) available at the
necessary scale? Damage assessments can be
designed to find out more about capabilities.
Determining whether the resources are available
should be part of assessments of needs and
resources, treated in more detail in PCR Tool 4.
How can we learn?
The four assessment methods highlighted in the
above section are examples of learning methods
specifically designed to provide answers to
particular questions. We will explain those in
more detail in the text below and PCR Tool 4:
Assessment of Reconstruction Needs and
Resources. There are, however, a number of
additional participatory tools available that can
help to answer the above questions. Many planning
tools start off with developing an understanding
of a particular set of problems, which is a
learning process. Practical Action South Asia has
summarised participatory learning and action tools
in a technical brief. Community Action Planning
(CAP) can also be a good learning experience
for participants; this is further described in the
Community Planning Website. For more detail, see
the resources at the end of this tool and PCR Tool
7: Planning with the People.
1 Vulnerability Assessment
Disasters do not result from hazards alone, but
from the impact of those hazards on communities
that are vulnerable and poorly prepared. Disasters
are not inevitable and communities are not
helpless. Action can be taken to build resilience
to hazards and strengthen capacity to adapt to
change. Practical Action has developed “from
Vulnerability to Resilience (V2R)”, a framework
to analyse vulnerability and plan for building
community resilience to guide this action.
Vulnerability is multi-dimensional; its roots may
lie in weak livelihoods; hazards and stresses;
future uncertainty (i.e. related to climate change);
or a poor governance environment. In the case
of the Alto Mayo, highlighted at the start of
this tool, livelihoods had weakened, governance
was worsening, and there were known risks
of earthquakes and floods in the region. More
recently, in Haiti, poverty played a major role, in
exacerbating vulnerability to the earthquake. As
72.1% of the population lived on less than $2 per
day, people could not afford to employ qualified
labour and built houses with poor quality materials.
2 Structural Damage Assessment
The experience of previous disasters has taught us
that buildings can be at risk of damage or collapse
for a number of reasons, the most common of
which are listed in the box below. These are the
kind of issues a damage assessment team will have
to look out for in particular.
Damage assessments should be done in teams
which include representatives of all stakeholders
involved in reconstruction. It is particularly
important for communities to identify their
indigenous knowledge on disaster mitigation. When
it comes to building houses, the starting point is
to learn from how people are building already, and
how they are incorporating specific details to reduce
disaster risk. Where livelihoods are concerned, it
is important to understand the strategies already
in place to cope with other types of crisis. Ideally,
such assessments are to be done by practitioners
visiting settlements, making observations and
having discussions with local builders and small
groups or representatives of different categories
of inhabitants. Their conclusions then need to
be presented and discussed with larger groups of
Where disasters have affected thousands of
people and large reconstruction programmes are
photo © Tek Sapkota, Practical Action Nepal
In such contexts where governance is weak, there
are no real mechanisms to enforce building codes.
In cities like Port-au-Prince, where many were
housed in poor and densely-packed shantytowns
and badly-constructed dwellings, the devastation
has been great and the death toll heavy. All of these
dimensions of vulnerability need to be analysed,
together with communities. The V2R framework
provides the background information and various
methods for doing so.
Within the context of this toolkit, the livelihood
dimension of increasing resilience is largely dealt
with in Tool 6: Integrating Livelihoods, and the
dimension of involving communities in governance
is dealt with particularly in Tool 7: Planning
with the People. In this tool, we will therefore
concentrate on the assessment of hazards and
stresses, and to an extent on how these might
change in future. Vulnerable people often lack a
good understanding of hazards and their associated
risks. If they are to become more resilient, they will
have to develop capacity to analyse and understand
the hazards and stresses that affect their lives.
To achieve this, agencies can work directly with
communities to carry out a systematic vulnerability
analysis, and/or train community leaders to
facilitate community analysis. An analysis of
vulnerabilities can be done in a participatory
way, and can help to identify households within a
community and those most in need of support. For
an example of Participatory Vulnerability Analysis,
see case study 3 in the section Applications.
House destroyed by flooding of the Jugedi river near Khetbari,
Nepal in September 2006. In a disaster risk assessment, this
location would be classified as high-risk.
envisaged, it becomes very difficult to ensure
everybody participates equally in assessments.
Reconstruction agencies may have to work with
representatives of the communities, but they should
ensure that this does not lead to the exclusion of
the opinions of vulnerable groups. Information and
communication technologies (ICTs) can help to
inform and involve larger numbers; e.g. providing
people with cameras or video equipment and a bit
of training can enable them to produce their own
stories of damage and vulnerability. Communication
is discussed in more detail in PCR Tool 9:
Communicating Better Building.
The aim of damage assessments is to find our
why some houses were badly damaged and others
less so; the box to the right indicates some of the
issues that need to be investigated. In addition,
it is important to learn what actions local people
had already taken to protect themselves from
disaster risks. Were these effective? If not, why
not? How could they be improved? Would these
improvements be suitable for the community to
carry out themselves during reconstruction, or
would additional support be needed?
If field workers involved know about similar
scenarios in other locations, where particular
improvements have proven to work well, they
can bring those into the discussion as examples.
However, field workers should take care not to take
on the role of experts and perhaps manipulate
communities towards certain solutions. The
purpose of the damage assessment is to learn why
the damage was caused, and how this could be
mitigated, not to immediately decide how houses
should be reconstructed. Field workers should in
particular be cautious not to suggest too many
technologies that are alien to local residents and
builders, since these would require additional
training, might require materials from elsewhere,
and quite possibly extra funding.
If particular types of houses survived the
disaster well, these can become the model for
future reconstruction. Alternatively, if it is observed
Analysis of Hazards and Stresses
The following method is a guide for the analysis. Since each context is different and specific, you may add other
questions that appear useful to ask. Ask people to tell you stories on past events, how hazards impacted on their lives,
and how they coped. Try to build up a rich and detailed picture of the kinds of hazards and stresses people faced,
which groups were most affected, and what opportunities exist for strengthening resilience.
Identify what different hazards and stresses have affected the community or particular groups in living memory,
both on a regular basis and one-offs.
Prioritise the different hazards, e.g. according to severity, numbers affected or frequency.
Further explore the prioritised hazards with the following questions and tools:
What is the typical frequency and duration of this hazard; has it changed over time?
Are there any warning signs that a hazard event is likely to occur; are there any early warning systems?
Are there any underlying causes of the hazards or stresses and does the community understand them, or
how to address them?
Which groups within the community are most affected and how?
Which communal or individual assets are affected and how?
How do different groups typically respond immediately after the hazard occurs (are there contingency
plans, safe areas, emergency resources, response organisations etc.)?
What particular long term coping strategies do these people (and particularly vulnerable groups) use to
recover from the hazard impact?
Based on the issues raised, what opportunities and capacities are available, or could be strengthened to
help people cope and recover when hazards and stresses occur?
Suggested tools to use: group discussion; hazard mapping; story telling; EMMA toolkit (to analyse changes to
Common Reasons for Disaster Damage to Buildings
The technology chosen is inherently unsafe for the type of hazards that may occur in a given location. An
example of this is the traditional house of adobe, stone or brick walls, with wooden poles lying across those
supporting cane and mats covered by a heavy layer of earth. Such roofs tend to move during earthquakes,
pushing the walls outwards, caving in on inhabitants. Earth walls are also very vulnerable to rain and flooding.
The building is poorly designed, e.g. with windows and doors close to corners or wall intersections weakening
the walls, or with irregular shapes which reduce its resistance to earthquakes, or with large roof overhangs or
verandas that can be ripped apart by strong winds.
Buildings are poorly located, e.g. on steep slopes with a risk of land slides, on alluvial plains at risk of flooding
or liquefaction, or on sites particularly exposed to strong winds (addressed in more detail under disaster risk
Protection provided is insufficient to resist hazards of more than a medium magnitude, e.g. houses are built on
plinths or columns, but only to a height that saves them from minor floods.
The quality of work is not good enough. This happens when people build themselves where a skilled craftsman
is needed, or use novel technologies they do not properly understand. Reinforced concrete frames can provide
resistance to several disasters, but they are often poorly erected, and therefore regularly collapse.
The protection provided is stretched beyond its specification, e.g. people add floors to a building, or make
changes to its design.
Residents are unable to maintain their houses adequately, which can cause components to weaken e.g. through
humidity or insect attack.
that although some of the houses performed
better than others there is still scope to improve
them, then a modified design can be developed.
However, if such modifications add a lot to the
cost of construction, their take-up is likely to be
low – certainly when limited external assistance is
available – even if people recognise them as being
more disaster-resistant. Fortunately, though, it is
often possible to make houses a lot more disasterresistant without adding more than 10% to the
original construction budget. This works best if any
such change is accompanied by awareness raising
initiatives, demonstration, and training on how to
build and maintain such houses.