Citation
Aid and Recovery in Post-Earthquake Nepal

Material Information

Title:
Aid and Recovery in Post-Earthquake Nepal Independent impacts and recovery monitoring phase 3, quantitative survey, September 2016
Creator:
The Asia Foundation ( Author, Primary )
UK aid ( contributor )
Swiss Development Cooperation ( contributor )
Place of Publication:
San Francisco, CA
Publisher:
The Asia Foundation
Publication Date:
Copyright Date:
© 2016, The Asia Foundation
Language:
English

Subjects

Subjects / Keywords:
एशिया -- नेपाल
Asie -- Népal
Educational impacts ( SWAY )
Education -- Impact ( LCSH )
शैक्षिक प्रभाव ( SWAY )
Displacement ( SWAY )
Refugees ( LCSH )
विस्थापन ( SWAY )
Reconstruction and development ( SWAY )
Nepal -- Economic development ( LCSH )
Nepal -- Repair and reconstruction ( LCSH )
पुन:निर्माण तथा विकास ( SWAY )
Subsidies and compensation ( SWAY )
Subsidies ( LCSH )
Wages ( LCSH )
अनुदान र क्षतिपूर्ति ( SWAY )
Grievances ( SWAY )
Grievance arbitration ( LCSH )
Grievance procedures ( LCSH )
गुनासोहरु ( SWAY )
Economic impacts ( SWAY )
Economic impact analysis ( LCSH )
आर्थिक प्रभाव ( SWAY )
Genre:
NGO Report ( SWAY )
Temporal Coverage:
20150429 - 20160109
Spatial Coverage:
Asia -- Nepal
Coordinates:
28 x 84

Notes

General Note:
Funded by GCRF (Global Challenges Research Fund) through AHRC (Arts and Humanities Research Council), Grant number AH/P003648/1, as "After the Earth's Violent Sway: the tangible and intangible legacies of a natural disaster", Dr. Michael Hutt, Principal Investigator.

Record Information

Source Institution:
SOAS University of London
Rights Management:
All applicable rights reserved by the source institution and holding location.

Downloads

This item has the following downloads:


Full Text
Independent Impacts and
Recovery Monitoring Phase 3
September 2016
Synthesis Report
â–  â–  -
"3- ^-Tr_ — 1 . □ 5 • ■ "i J '■ " -->rr- J J
——-~TZ ^S~T"1 "** "*
â– i" j 35* a *
■‘v ■

1 5
i f
a 1 k.
its ■ • u ’ -~t . ■. —- ■ -■j|
uKaid
from the British people
Sdiweizerischa EidgeruMseriEctiaft
ConfeddraHon guigso
ConfetJarazione Svizzera
ConfedsraziLim svizra
The Asia Foundation
Faderal Department of Foreign Affairs FDFA
Swlw Agency for Development and Cooperation SDC
’sfpr ■*.'-TT? lit-fM Mfq'l 'I




Synthesis Repo
Independent Impacts and Recovery Monitoring Phase 3:
September 2016
Scfiwetwtiflche EidflenowenacMil
Conl'frMfflllon Euisse
Confedecazcine Svizzera
Conredorazium s*izra
The Asia Foundation
UKaid
from the British people
Federal Departmartl of Foreign Afiairs FDFA
Swiss Agency for Devetopnient and Cooperalton SDC


The Asia Foundation is a nonprofit international development organization committed to improving lives across a dynamic and developing
Asia. Informed by six decades of experience and deep local expertise, our work across the region addresses five overarching goals—
strengthen governance, empower women, expand economic opportunity, increase environmental resilience, and promote regional
cooperation. Headquartered in San Francisco, The Asia Foundation works through a network of offices in 18 Asian countries and in
Washington, DC.
Independent Impacts and Recovery Monitoring Phase 3
Synthesis Report
© The Asia Foundation
All rights reserved. No part of this book may be reproduced
without written permission from The Asia Foundation
The Asia Foundation
456 California Street, 9th Floor
San Francisco, CA U.S.A. 94104
www.asiafoundation.org
The project is funded by UK aid through the UK government and the Swiss Development Cooperation.
The views expressed in this report do not necessarily reflect the UK or the Swiss government’s official policies.
Cover photo: Nayan Pokharel, Binu Sharma
Design: Deddeaw Laosinchai


Aid and Recovery in Post-Earthquake Nepal
PREFACE
In the early weeks after the earthquakes of April and
May 2015, The Asia Foundation conducted a study
aimed at assessing its impacts on the ground and
understanding whether the emergency aid that was
flowing in to affected areas was helping people recover.
Using both quantitative and qualitative methods,
the initial study highlighted just how destructive the
earthquakes had been and the immense challenges
that would lie ahead. Since then, two further rounds
of mixed methods research have been conducted in the
same areas, allowing for a tracking of how recovery has
been occurring. The second round of research, which
involved fieldwork almost a year after the disasters,
highlighted new emerging issues. Borrowing had risen
massively and the reports discussed the potential for
the poor and marginalized to get stuck in a vicious debt
trap. Very few at that point had moved from temporary
shelters into more sturdy housing. It was clear that the
livelihoods of many people, in particular farmers, was
recovering very slowly. And tensions were brewing
related to a series of contentious damage assessments
and perceived mistargeting of aid.
This report presents findings from the third round
of research, conducted in September 2016 almost
eighteen months after the earthquakes. Because each
round of research takes place in the same areas, with
the same people interviewed where possible, the
series of studies provides insights into how people’s
experiences and perceptions are evolving over time.
Between the second and third round of fieldwork, the
process of distributing reconstruction cash grants to
those whose houses were destroyed or badly damaged,
and who were identified as beneficiaries during a new
round of assessments, began. This report provides
insights into this process and the impacts it has had.
It also looks, amongst other things, at overall progress
made with regards to reconstruction in the research
areas, the make-up of aid in the earthquake-affected
zone, and remaining needs. Further, the report dis-
cusses the roles and involvement of political parties
and other local leaders during reconstruction, changes
to social relations, protection issues and vulnerable
groups, impacts on the local economy and people’s
livelihoods, and the coping strategies people are using
and their effectiveness.
George Varughese, Ph.D.
Nepal Country Representative
The Asia Foundation
Patrick Barron, Ph.D.
Regional Director for Conflict & Development
The Asia Foundation


Aid and Recovery in Post-Earthquake Nepal
Acknowledgements
This report summarizes and synthesizes findings
from the third round of the Independent Im-
pacts and Recovery Monitoring (IRM) project
in post-earthquake Nepal. IRM consists of in-depth
qualitative fieldwork and large quantitative surveys
conducted in earthquake-affected areas. The synthesis
report was written by Lena Michaels, Sasiwan Ching-
chit, and Patrick Barron.
The qualitative work was conducted by researchers
from Democracy Resource Center Nepal (DRCN),
led by Sudip Pokharel and coordinated by Apurba
Khatiwada. Analysis of the data was done by Apurba
Khatiwada, Soyesh Lakhey, Amanda Manandhar
Gurung, Shekhar Parajulee, and Sudip Pokharel, who
co-authored the qualitative report with TAF contrib-
utors, Lena Michaels, and Sasiwan Chingchit. Special
thanks goes to the team of researchers for their dedi-
cation in the field: the lead researchers Anubhav Ajeet,
Chiran Manandhar, Ishwari Bhattarai, Nayan Pokhrel,
Shekhar Parajulee, and Subhash Lamichhane; and
the researchers Alok Pokharel, Anurag Devkota, Binu
Sharma, Janak Raj Sapkota, Punam Limbu, and Tanka
Gurung.
The survey was implemented by a team from Inter-
disciplinary Analysts (IDA) led by Sudhindra Sharma.
Other IDA staff who provided support to this survey
included Hiranya Baral, Bal Krishna Khadka, Chandra
KC, and Sandeep Thapa. The data was analyzed by
Anup Phayal, Jui Shrestha, and Patrick Barron.
A number of other people provided useful inputs
at various stages, including in the formation of the
research questions, finalization of the sample, and
analysis of the data. They included George Varughese,
Nandita Baruah, and James Sharrock. Deddeaw
Laosinchai designed the report’s layout.
The Asia Foundation wishes to express its appreci-
ation to the informants and survey respondents in
Bhaktapur, Dhading, Gorkha, Kathmandu, Lamjung,
Nuwakot, Okhaldhunga, Ramechhap, Sindhupalchok,
Solukhumbu, and Syangja.
The IRM project is funded by UK aid through the UK
government and the Swiss Development Cooperation,
with support from the UK Department for International
Development’s Programme Partnership Arrangement
with The Asia Foundation. Andy Murray (UK DFID)
and Pia Haenni (SDC) have managed the project from
the donor side, and have provided useful inputs at
every stage. Thanks also to Stefan Feurst (SDC) and
Craig Irwin (DFID) who took over management of the
project as the report was being produced. The views
here do not necessarily reflect the UK or the Swiss
governments’ official policies.
The IRM research is directed by Patrick Barron with
assistance from Sasiwan Chingchit. Lena Michaels
coordinates the project in Nepal with support from
The Asia Foundation-Nepal.
iv


Aid and Recovery in Post-Earthquake Nepal
Executive Summary
This synthesis report combines and summarizes
findings from the third wave of the Independent
Impacts and Recovery Monitoring for Account-
ability in Post-Earthquake Nepal (IRM), a longitudinal
mixed method research project designed to monitor
aid impacts and patterns of recovery in earthquake-af-
fected areas. The first round of research was conducted
in June 2015 and the second in February-March 2016.
Fielding of the third round was carried out in eleven
affected districts for the quantitative survey and in six
districts for the qualitative component in September
2016. Districts included those in four categories of
earthquake impact identified by the government’s
Post-Disaster Needs Assessment: severely hit districts
(those most affected), crisis hit districts (second high-
est impact category), hit with heavy losses districts
(third category), and a hit district (the least impacted
of those affected).
At the time IRM-2 was being conducted, the National
Reconstruction Agency (NRA) had only just begun
its work. Since then, the NRA has focused largely
on housing reconstruction, in particularly through
the Nepal Rural Housing Reconstruction Program
(RHRP), which is supported by a multi-donor fund.
Under the RHRP, which emphasizes owner-driven
reconstruction, cash grants of NPR 300,000 are
provided in three instalments to eligible beneficiaries
to aid them in building earthquake-resistant houses.
By September 2016, when IRM-3 was conducted, the
signing of beneficiary agreements and distribution of
the first installment of the reconstruction cash grant
was underway in 11 districts, including four visited in
the qualitative fieldwork and eight where the quanti-
tative survey was conducted. However, many people
had not yet received cash in hand and progress in
reconstruction remained slow. Since then, there has
been more progress with the RHRP. However, be-
cause these developments came after fieldwork was
conducted, they are not covered in this report. It is
within this context that IRM-3 presents a picture of
recovery, aid effectiveness, and coping strategies of
people in affected areas, one and a half years on from
the earthquakes.
Recovery
Housing and shelter. As of September 2016, more
people had moved back into permanent housing
but the number of people in temporary shelters was
still very high in severely hit districts. People from
marginalized groups were disproportionately likely
to still be living in shelters. Movements to and from
shelters were significant between all research rounds.
While many people left their temporary shelters to
move back into their houses, some also moved out
again, returning to temporary shelters or moving into
the homes of others after realizing that their house or
the land it was on remained unsafe. Those staying in
shelters had faced two monsoons and one winter since
the earthquakes by September 2016. Many shelters
were not suited for adverse weather causing people
to fall sick during the monsoon months.
Progress in the reconstruction of homes had been
slow. Of those whose house was impacted, most people
had done nothing to repair or rebuild. Alack of money
was the primary factor that prevented people from
starting to rebuild their houses. Larger sums were
needed to rebuild compared to pre-earthquake times
as construction costs had increased significantly.
Other commonly cited reasons were people waiting
to receive the reconstruction cash grant or a lack of
knowledge on procedures and technical requirements
of the cash grant program.
Livelihoods. Different livelihoods continued to
recover, with recovery more widespread in mid-2016
compared to early 2016. Very few have changed
livelihoods since the earthquakes. However, many
pre-existing hardships were exacerbated, pushing
those already poor further into poverty, especially
poor farmers and Dalits.
Food. The need for food in all affected districts de-
clined in IRM-3 compared to earlier research rounds
and food consumption has remained similar between
IRM-2 and IRM-3. Severely impacted districts and
more remote and rural areas reported the greatest
need for food.
V


Aid and Recovery in Post-Earthquake Nepal
Trauma and vulnerability. Many people were
still suffering psychologically from the earthquakes.
Extreme fear and startling while sleeping were the
most common psychological effects. The displaced
and those living in temporary shelters remained
among the most vulnerable groups. Landslides also
continued to be a common worry and risk factor.
Geological landslide assessments remained important
to assess risks and determine long-term resettlement
for the displaced. Inequality and prevalent forms of
exclusion and discrimination are negatively affecting
the recovery of marginalized groups, especially of
Dalits who stood out as a highly vulnerable group in
IRM-3. People in very remote areas were also facing
greater obstacles to accessing aid and rebuilding
their houses. Women, children, and the elderly faced
particular challenges and continued to be seen as
particularly vulnerable groups.
Aid delivery
Coverage of aid. The coverage of aid declined
massively between IRM-2 (February-March 2016)
and IRM-3 (September 2016) with only 15% receiving
aid now. The drop in the coverage of aid was true for
different types of assistance including relief, material
aid, and cash support.
The government remained the most prominent aid
provider, followed by INGOs and NGOs. The share of
people receiving aid from individual donors declined
significantly. Government and non-government
providers offered different types of assistance. The
government focused on the distribution of cash grants
while I/NGOs mostly provided ‘soft’ forms of aid
through trainings, awareness raising, and technical
assistance.
Needs. Needs have changed over time but the drop
in aid coverage did not correspond with any declining
demand for aid. On the contrary, as time passes,
the gap between needs and aid provided seems to
be increasing. This was exacerbated by the fact that
there seemed to be no shared understanding and little
coordination at the local level to identify and prioritize
needs. Cash and construction materials remained the
most widely cited need.
Housing reconstruction program. Overall,
perceptions of the RHRP were not favorable. People
were more satisfied with the agreement process
than with the assessment to determine eligibility or
with access to the grant. Dissatisfaction was highest
over the size of the cash grant. While the grant was
intended as incentive to build earthquake-resilient
buildings, not to fully cover construction costs, many
were dissatisfied with the amount as they thought it
was insufficient. Estimates of construction costs show
that the grant will likely only cover a small share of the
costs. The government has made provisions to provide
soft loans to help with housing reconstruction but this
had not happened in practice at the time of research.
Few planned to use the first installment of the cash
grant for the intended purpose. Limited technical
assistance was provided at the time of research. Where
deployed engineers were present, they were often
inactive and waiting for cash grants to be distributed
and rebuilding to begin.
Coordination. Coordination was generally weak at
the local level, both between different government of-
fices and between government and non-governmental
organizations. Overlap of or confusion over respective
responsibilities hindered effective coordination and
affected the reconstruction process.
Communication and satisfaction with aid
providers. Satisfaction with every aid provider
decreased significantly between IRM-2 and IRM-3.
People said dissatisfaction with I/NGOs was rising be-
cause of their alleged disregard of people’s needs when
designing and implementing programs. Increasing
dissatisfaction with the government and political par-
ties was largely due to delays in the provision of cash
grants, unclear policies and information, and delays
in addressing complaints. Perceptions of the fairness
of the distribution of aid by VDCs or municipalities
also markedly declined. Fewer people than before
thought that everyone could get aid according to their
needs than in the past. The most common source for
information about aid were neighbors, radio, the VDC
office, and Ward Citizen Forums. However, levels of
satisfaction with communication with aid providers
were low. People did not feel that they could communi-
cate well with aid providers, especially those removed
from the local level.
Coping strategies
Borrowing. The proportion of people borrowing
in IRM-3 remained similar to IRM-2 but was much
higher than in IRM-1. Amounts borrowed also
increased since IRM-1. Average monthly interest rates
for many sources, especially informal ones, increased
slightly since IRM-2 suggesting a growing demand
for credit. Livelihoods, food, and rebuilding houses
were the main reasons for borrowing. Shelter-related
borrowing (temporary shelter, rebuilding houses,
improving temporary shelters) was concentrated in
the severely hit districts. There were indications that
borrowing is likely to increase over time. Many said
they planned to borrow in the near future to cope with
the impacts of the earthquakes.
Borrowing was higher among already vulnerable
groups. People in remote and rural areas, in severely
vi


Aid and Recovery in Post-Earthquake Nepal
hit districts, low caste individuals, those with lower in-
come, and those in temporary shelters reported higher
rates of borrowing. These groups were borrowing more
frequently, at higher interest rates, and were more
likely to say that they planned to borrow in the near
future. Repeat borrowers have been less likely to see
livelihoods recovery or to move home and have been
more likely to see reductions in food consumption.
Rising debts were a worry for many households and
the risks of debt traps were increasing. Dalits faced
particular difficulties accessing credit especially from
formal sources.
Migration and remittances. People in severely
hit districts were slightly more likely to have moved
since the earthquakes. The most common reason for
migration were lack of shelter, lack of livelihoods,
and landslides. The volume of remittances received
remained largely the same. The share of household
identifying remittances as a main income source grew
over the three research rounds but the number of
those reporting to have received them has declined.
However, many households said they were planning
to send at least one family member abroad for work
if they faced difficulties paying for the reconstruction
of their houses and to pay back loans.
Politics, social cohesion, and conflict
Political parties. With the decline in emergency
relief, and the increasing focus on reconstruction,
the formal influence of political parties over the coor-
dination of assistance reduced. Yet, political parties’
role in local governance remained the same. And in
many areas, political party representatives were infor-
mally involved in the cash grant agreement process,
initially by supporting local obstructions and later by
negotiating agreements to resume the process and
by facilitating communication between government
offices and local communities. Dissatisfaction with
political parties, however, was high but this did not
lead to changes in which political party people said
they were supporting.
Social relations, security and crime. Most peo-
ple felt safe. There was no change in the proportion
of people feeling safe between the last two rounds
and very few people reported any violent incidents in
their community. Social cohesion has also generally
been strong since the earthquakes and social relations
remained largely unchanged between IRM-2 and
IRM-3. Conflicts and tensions continued where local
disagreements over displacement and resettlement
had not been addressed. Caste-based discrimination
was also common. Water shortages seemed to aggra-
vate tensions. Conflicts may escalate in the future if
tensions related to resettlement, water shortages,
and caste-based discrimination remain unaddressed.
Further, frustrations of earthquake victims over the
slow pace of reconstruction and policy changes may
rise if assistance is delayed further.
Focus areas and recommendations
The report presents independent recommendations
which are not necessarily those of the UK or Swiss
governments:
1) Shelter and housing reconstruction
• Communicate information on government cash
grant procedures more quickly and clearly to local
government offices and citizens. Local stakehold-
ers, who are close to affected communities, should
be utilized more for sharing information.
• ollect information on challenges related to ac-
cessing the grants after agreements have been
signed, and on the number of people who have
yet to withdraw the grant from bank accounts.
• Technical assistance during reconstruction needs
to be more widely available.
• Strengthen coordination mechanisms and infor-
mation flows between the NRA and government
line ministries in Kathmandu, districts headquar-
ters, and the local level. Roles and responsibili-
ties of different bodies need to be more clearly
defined.
• Develop plans for the clear transfer of responsibil-
ities related to reconstruction and recovery work
to new local bodies after local body restructuring.
• Improve the quality of shelters for the medium-
term and prioritize programs to mitigate the
consequences of staying in temporary shelter
(targeted health support and medicine, temporary
water and sanitation facilities, women’s security).
• Complete assessments to determine whether
people can return to and rebuild on land deemed
to be at risk. Clearly communicate the findings
of such assessments to local stakeholders and
affected households.
• Generate policy for supporting the permanent
resettlement of displaced households unable to
return to their land.
2) Debt and borrowing
• Expand soft loan programs, strengthen commu-
nication about them, and ensure they reach those
in remote areas and marginalized groups.
• Ensure better awareness of government low in-
terest loans in particular and make these more
widely available. Central-level loan policies may
need to be revised to ensure better access for
those in need of credit.
vii


Aid and Recovery in Post-Earthquake Nepal
3) Needs beyond reconstruction
• Strengthen communication channels for local
communities to express their needs.
• Track long-term psychosocial impacts of the
earthquakes and their implications for recovery
and expand psychosocial support for earth-
quake-affected communities.
• Continue to provide livelihood support to help
generate incomes for poor households, especially
for farmers.
4) Making sure the marginalized do not
get left behind
• Pay more attention to the specific challenges
of vulnerable groups to facilitate special assis-
tance that enhances their ability to recover. This
includes the need to develop a greater under-
standing of who is vulnerable in local areas and
the factors preventing vulnerable groups from
recovering.
• Targeted aid should be context-sensitive; this
means local communities need to be informed of
and involved in the development and implemen-
tation of targeted aid programs to avoid conflict.
viii


Aid and Recovery in Post-Earthquake Nepal
LIST OF ACRONYMS
CBS Central Bureau of Statistics
CGI Corrugated Galvanized Iron
CL-PIU Central Level Programme Implementation Unit
CPN-UML Communist Party of Nepal (Unified Marxist Leninist)
DCC District Coordination Committee
DDC District Development Committee
DDRC District Disaster Relief Committee
DFID UK Department for International Development
DRCN Democracy Resource Center Nepal
DL-PIU District Level Programme Implementation Unit
DRCN Democracy Resource Center Nepal
DUDBC Department of Urban Development and Building Construction
IDA Interdisciplinary Analysts
INGO International non-governmental organization
IRM Independent Impacts and Recovery Monitoring for Accountability in Post-Earthquake Nepal project
IRM-1 First round of the IRM study (June 2015)
IRM-2 Second round of the IRM study (February-March 2016)
IRM-3 Third round of the IRM study (September 2016)
MoUD Ministry of Urban Development
MP Member of Parliament
NGO Non-governmental organization
NPR Nepali Rupees
NRA National Reconstruction Authority
PDNA Post-Disaster Needs Assessment
RHRP Rural Housing Reconstruction Program
RPP-N Rajastriya Prajatantra Party Nepal
SDC Swiss Agency for Development and Cooperation
TAF The Asia Foundation
UCPN-MC Communist Party of Nepal (Maoist-Centre)
UN United Nations
VDC Village Development Committee
WCF Ward Citizen Forum
ix


Aid and Recovery in Post-Earthquake Nepal
TABLE OF CONTENTS
PREFACE III
ACKNOWLEDGEMENTS IV
EXECUTIVE SUMMARY V
LIST OF ACRONYMS IX
LIST OF FIGURES XI
LIST OF TABLES XIII
LIST OF CASE STUDIES XV
Chapter 1. Introduction 1
1.1 Background 1
1.2 Methodology 2
1.3 Contextual changes since March 2016 5
1.4 Report structure 6
Chapter 2. Recovery 7
2.1 Housing and shelter 7
2.2 Infrastructure and service delivery 17
2.3 Livelihoods 19
2.4 Food 24
2.6 Trauma and vulnerabilities 26
Chapter 3. Aid and Housing Reconstruction Cash Grants 29
3.1 Aid delivery 29
3.2 Needs 36
3.3 Housing reconstruction cash grants 38
3.4 Coordination of aid 45
3.5 Communication and satisfaction 46
Chapter 4. Coping Strategies 53
4.1 Borrowing 53
4.2 Sale of assets 63
4.3 Migration and remittances 64
Chapter 5. Politics, Social Cohesion, and Conflict 67
5.1 Roles of political parties in the provision of aid 67
5.2 Satisfaction with political parties
and future vote preferences 69
5.3 Security and crime 72
5.4 Trust and social cohesion 74
5.5 Potential sources of conflict 76
Chapter 6. Conclusions and Recommendations 77
6.1 Overview of conclusions 77
6.2 Key focus areas and recommendations 79
Annex A. The current status of reconstruction 83
X


Aid and Recovery in Post-Earthquake Nepal
LIST OF FIGURES
Figure 1.1: Analytic framework 2
Figure 2.1: Where people were/are living (IRM-1, IRM-2, IRM-3, weighted) 7
Figure 2.2: Where people are living - by district impact (IRM-3, weighted) 8
Figure 2.3: Share of people living in different types of shelters (IRM-2, IRM-3, weighted) 9
Figure 2.4: Share saying they have the services provided by VDC/municipality 17
(IRM-1, IRM-2, IRM-3, weighted)
Figure 2.5: Changes in quality of services (IRM-1, IRM-2, IRM-3, weighted) 18
Figure 2.6: Satisfaction with public services (IRM-1, IRM-2, IRM-3, weighted) 19
Figure 2.7: Share of people within each income source whose income from that 20
source has improved in the last three months - by source of income
(IRM-2, IRM-3, weighted)
Figure 2,8: Food as a top immediate need and three month need (IRM-1, IRM-2, 24
IRM-3, weighted)
Figure 2,9: Food as a top immediate need and three month need - by district impact 25
(IRM-3, weighted)
Figure 3.1: Proportion of people receiving aid - by district impact (IRM-1, IRM-2, 30
IRM-3, weighted)
Figure 3.2: Proportions of people receiving and not receiving cash who moved 32
from shelter to home - by government vs. non-government cash
(IRM-1, IRM-2, IRM-3 household panel, unweighted)
Figure 3.3: Sources of aid amongst those who received aid (IRM-1, IRM-2, IRM-3, weighted) 33
Figure 3.4: Proportion who received aid - by caste (IRM-1, IRM-2, IRM-3, weighted) 35
Figure 3.5: Changes in priority needs - IRM-1 and IRM-2 current and future needs, 37
IRM-3 current need (IRM-1, IRM-2, IRM-3, weighted)
Figure 3.6: Plans for use of RHRP grant amongst those declared eligible (IRM-3, weighted) 41
Figure 3.7: Citizens’ perception of the government’s cash grant program (number of 43
wards - qualitative research)
Figure 3.8: Sources of information on aid (IRM-3, weighted) 47
Figure 3.9: Satisfaction with communication with aid providers (IRM-3, weighted) 47
Figure 3.10: Satisfaction with how aid providers communicate about aid - by aid 48
provider (IRM-3, weighted)
Figure 3.11: Opinions on whether everyone can get aid according to their needs 51
(IRM-2, IRM-3 household panel, unweighted)
Figure 3.12: Groups who are unable to get aid equally according to their needs among 52
those who disagree that everyone can get aid equally (IRM-3, weighted)
Figure 4.1: Sources of borrowing among those who borrowed (IRM-1, IRM-2, IRM-3, weighted) 55
Figure 4.2: Reasons for borrowing, share of those borrowing (IRM-2, IRM-3, weighted) 56
Figure 4.3: Share of people who have borrowed - by housing damage (IRM-1, IRM-2, 57
IRM-3, weighted)
Figure 4.4: Share of people who have borrowed - by rural/urban and remoteness 58
(IRM-1, IRM-2, IRM-3, weighted)
Figure 4.5: Share of people who have borrowed since the end of winter - by where 58
people are living (IRM-3, weighted)
xi


Aid and Recovery in Post-Earthquake Nepal
Figure 4.6: Frequency of borrowing across the last two surveys - by income 60 (IRM-2, IRM-3 household panel, unweighted)
Figure 4,7: Figure 4.8: Unsuccessful borrowers - by where people are living (IRM-3, weighted) 61 Remittances as a main income source - by district impact (IRM-1, IRM-2, 64 IRM-3, weighted)
Figure 4.9: Reasons for migration (IRM-2, IRM-3, weighted) 65
Figure 5.1: Figure 5.2: Voting preference for next election (IRM-1, IRM-2, IRM-3, weighted) 70 Perceptions of safety in the community - by district impact (IRM-1, 72 IRM-2, IRM-3, weighted)
Figure 5.3: Perceptions of safety in the community - by where people are living 73 (IRM-2, IRM-3, weighted)
xii


Aid and Recovery in Post-Earthquake Nepal
LIST OF TABLES
Table 2.1: Share of people who were in shelter (IRM-1) to their own house (IRM-3) - by 10
district impact and district (IRM-1, IRM-2, IRM-3 household panel, unweighted)
Table 2.2: Share of people who moved from shelter (IRM-1) to their own house 10
(IRM-3) - by what people have done to their house (IRM-1, IRM-2, IRM-3
household panel, unweighted)
Table 2.3: Actions to repair or rebuild houses amongst those whose house 13
was impacted - by district impact, district, rural/urban and remoteness
(IRM-3, weighted)
Table 2.4: Actions to repair or rebuild houses amongst those whose house was 13
impacted - by gender, caste, income and disability (IRM-3, weighted)
Table 2.5: Reasons for stopping repairing or not building a house - by district impact, 14
district, remoteness, rural/urban and income (IRM-3, weighted)
Table 2.6: Cost of construction materials (IRM-3, weighted) 15
Table 3.1: Proportion of people receiving aid - by district impact and district 30
(IRM-1, IRM-2, IRM-3, weighted)
Table 3.2: Proportion of people receiving different types of aid (IRM-1, IRM-2, 31
IRM-3, weighted)
Table 3.3: Type of aid provided - by source (IRM-3, weighted) 33
Table 3.4: Top five current needs - by district impact and district (IRM-3, weighted) 36
Table 3.5: Housing classification in most recent damage assessment - by self-reported 39
housing damage (IRM-3, weighted)
Table 3.6: Satisfaction with official damage classification - by district and district 39
impact (IRM-2, IRM-3 household panel, unweighted)
Table 3.7: Eligibility for RHRP grant - by housing damage classification (IRM-3, weighted) 40
Table 3.8: Plans for use of RHRP grant amongst those declared eligible - by district 42
impact and district (IRM-3, weighted)
Table 3.9: Share of people satisfied with aid providers on how information on aid 49
was given - by whether people think communication was good or bad
(IRM-3, weighted)
Table 3.10: Proportion satisfied with aid provider (IRM-1, IRM-2, IRM-3, weighted) 49
Table 3.11: Share of people who agree that VDC/municipalities are distributing aid 51
fairly - by district impact and district (IRM-1, IRM-2, IRM-3 household
panel, unweighted)
Table 4.1: Share of people who borrowed money (IRM-1, IRM-2, IRM-3, weighted) 53
Table 4.2: Average borrowing in NPR - by district impact and district (IRM-1, IRM-2, 54
IRM-3, weighted)
Table 4.3: Mean of self-reported amount (in thousand NPR) borrowed from different 54
sources in the three survey waves (IRM-1, IRM-2, IRM-3 household panel,
unweighted)
Table 4.4: Average borrowing in NPR - by sources (IRM-1, IRM-2, IRM-3, weighted) 55
Table 4.5: Share of people who borrowed money - by caste, occupation and income 59
(IRM-3, weighted)
Table 4.6: Share of people who borrowed in both IRM-2 and IRM-3 - by district 59
impact, district, rural/urban and remoteness (IRM-2, IRM-3 household
panel, unweighted)
xiii


Aid and Recovery in Post-Earthquake Nepal
Table 4.7: Mean reported interest rates - by district impact, district, rural/urban and 60 remoteness (IRM-3, weighted)
Table 4.8: Table 4.9: Reasons for unsuccessful borrowing - by income and caste (IRM-3, weighted) 61 Changes in remittances since the earthquakes - by income and disability 65 (IRM-3, weighted)
Table 5.1: Table 5.2: Future voting preferences - by past voting behavior (IRM-3, weighted) 70 Proportion of people reporting violence in their community - by district 73 impact, district and remoteness (IRM-1, IRM-2, IRM-3, weighted)
Table A.i: Progress of private house reconstruction and cash grant distribution in the 84 qualitative research areas
xiv


Aid and Recovery in Post-Earthquake Nepal
LIST OF CASE STUDIES
Case Study 2.1: High transportation costs during the monsoon affect reconstruction 16
Case Study 2.2: Masonry continues to gain in Solukhumbu 21
Case Study 2.3: A displaced Dalit is struggling to resume farming 22
Case Study 3.1: Confusion over responding to complaints in Okhaldhunga 45
Case Study 4.1: A migrant laborer borrowing to rebuild 62
Case Study 4.2: Young entrepreneurs find new opportunities at home 66
XV




Aid and Recovery in Post-Earthquake Nepal
Photo: Chiran Manandhar
1.1 Background
This report assesses conditions on the ground in
earthquake-affected areas of Nepal in September 2016.
Combining quantitative and qualitative information,
it assesses the extent to which recovery is taking place
almost 18 months after the earthquakes.
The Independent Impacts and Recovery Monitoring
for Accountability in Post-Earthquake Nepal (IRM)
project tracks evolving conditions and needs in areas
of Nepal that were affected by the massive earthquakes
of April and May 2015. Using both quantitative survey-
ing and in-depth qualitative fieldwork, IRM involves
revisiting areas and people at roughly six month
intervals to assess current conditions and how they
are changing. Because data collection and research is
conducted in the same areas in each round, with many
of the same people interviewed, IRM allows for an
assessment of how conditions and needs are changing
over time and of the roles that aid is playing—positive
and negative—in shaping recovery patterns.
The pace of recovery, and the experiences of different
population groups, will be determined by the level of
earthquake impacts, the aid response, the coping strat-
egies employed by affected households and communi-
ties, and the political and economic context in which
the recovery is taking place. IRM focuses on each of
these issues at the local level to assess the extent to
which recovery is taking place, how this varies between
groups and areas, and the causes of differences in the
degree and nature of recovery (Figure 1.1).
This report provides findings from the third phase of
research (referred to as IRM-3). It combines findings
from quantitative and qualitative research.1 1 The report
provides data and analysis on the situation in Septem-
ber 2016, almost a-year-and-a-half after the initial
earthquakes, comparing the data with that collected
in the two past rounds: IRM-1 conducted in June 2015
and IRM-2 in February-March 2016.2 A fourth wave of
surveying and fieldwork is currently being conducted.
1 Reports, published in parallel, outline in greater depth findings
and analysis from the qualitative and survey research. The Asia
Foundation and Democracy Resource Center Nepal (2017). Aid
and Recovery in Post-Earthquake Nepal: Independent Impacts
and Recovery Monitoring Phase 3 - Qualitative Field Moni-
toring (September 2016). Kathmandu and Bangkok: The Asia
Foundation; The Asia Foundation (2017). Aid and Recovery in
Post-Earthquake Nepal: Independent Impacts and Recovery
Monitoring Phase 3 - Quantitative Survey (September 2016).
Kathmandu and Bangkok: The Asia Foundation.
2 Reports from previous rounds can be accessed at: http://
asiafoundation.org/tag/independent-impacts-and-recovery-
monitoring-nepal /
1


Introduction
Figure 1.1: Analytic framework
r i
Impacts of disaster
k J
Economy &
Livelihoods
Social Relations
& Violence
r
Degree &
Nature of recovery
L J
r 'I
Impacts of aid
k J
Protection &
Vulnerability
Politics &
Leadership
1.2 Methodology
The IRM-3 survey involved face-to-face interviews
with 4,855 respondents (plus surveys with 305 ward
leaders). These were conducted in 11 districts, all of
which were covered in the IRM-1 and IRM-2 surveys
(Map 1.1).3 * 3 IRM is set up as a panel survey meaning
that, where possible, the same people are interviewed
in each round. Respondents in IRM-1 were selected
using stratified randomized sampling. Subsequent
rounds of surveying sought to re-interview the same
people to allow for an assessment of changes over time.
Because the survey respondents are the same people,
we can be confident that any changes we find in survey
answers relate to changes on the ground rather than
to the make-up of the sample. The vast majority of
people interviewed in the IRM-3 survey (4,446 out
of the 4,855) had also been interviewed in IRM-2.
A smaller number of these people (1,470) were also
interviewed in IRM-1.4 In some places in the report,
we use the data that includes only people interviewed
in multiple rounds (referred to as the household panel
dataset). In other analyses, we use the full datasets
from IRM-1, IRM-2, and IRM-3.5 The IRM-3 survey
was deliberately designed to mirror the IRM-1 and
IRM-2 instruments, with many of the questions re-
maining the same. This allows for direct assessment
to be made of changes over time.
Data collection took place in districts that were strat-
ified using the categories of earthquake impact from
the Government’s Post-Disaster Needs Assessment
(PDNA): Nuwakot, Sindhupalchowk, Ramechhap,
Gorkha, and Dhading (severely hit); Bhaktapur,
Okhaldhunga, and Kathmandu (crisis hit); Solukhum-
bu and Lamjung (hit with heavy losses); and Syangja
(hit). Severely hit districts are the most affected
districts, followed by crisis hit districts, then hit with
heavy losses districts, and then hit districts.
3 The IRM-1 survey was conducted in 14 districts. Three of these
districts were dropped for IRM-2 and IRM-3. IRM-1 was con-
ducted before the government’s Post-Disaster Needs Assessment
(PDNA) was released and selection of districts was made from the
26 districts initially deemed affected by the government. Three of
the selected districts (Manang, Khotang, and Dang) surveyed in
IRM-i were subsequently not included in the PDNA’s classification
of earthquake-impacted districts. As such, they were not part of
the sample for the IRM-2 and IRM-3 surveys.
4 This is primarily because the sampling strategy changed after
IRM-i with three districts dropped and new wards selected in the
remaining 11 districts.
5 For a fuller discussion of the survey methodology, and changes in
approach over time, see the IRM-3 survey report.
2


Aid and Recovery in Post-Earthquake Nepal
Map 1.1: Locations of surveyed districts
The qualitative research involved teams conducting
interviews, focus group discussions, and participant
observation in six districts spread across different
earthquake impact categories: Sindhupalchowk,
Gorkha, and Ramechhap (severely hit); Okhaldhunga
(crisis hit), Solukhumbu (hit with heavy losses), and
Syangja (hit) (Map 1.2). Research teams visited 16
village development committees (VDCs) and two
municipalities, with two wards studied in each.
Research took three-four days per VDC and was
supplemented by interviews in district capitals.
Sampling of locations was done at three levels—
district, VDC, and ward—to maximize variation in
two factors that were predicted to affect the nature
and speed of recovery: the degree of impact of the
earthquake; and the degree of remoteness.
Map 1.2: Locations of qualitative research
3


Introduction
Photo: Chiran Manandhar
The methodology for both components of the research
was developed to ensure to the greatest degree possible
that findings accurately reflect conditions and views
in earthquake-affected areas. A few relevant details
regarding the methodology and its limitations should
be noted.
Timing of research
IRM-3 fieldwork was conducted from late August
until September 2016. During this period, large-scale
government reconstruction policies and schemes were
beginning to be rolled out (see Annex A). As such, the
report does not evaluate policies or aid provided after
September 2016. The fourth round of IRM will capture
more information on those developments.
Confidence in findings
The quantitative survey is representative of all people
in the eleven districts studied. A careful sampling
strategy—at the Village Development Committee
(VDC), ward, household, and individual levels—was
developed and employed. Stratified random sampling,
along with weighting of the data, means that we can
be sure with a high degree of confidence that what
we find holds true for the wider population living in
earthquake-affected districts. The margin of error
across the whole dataset is +/- 1.4% at a 95 percent
confidence level. The sample size is at least 350 for
each district allowing for a margin of error of +/- 5.2%
for district-disaggregated analyses. Where we break
down the survey population by impact, demographic,
or other variables (for example, comparing the
opinions of men and women or patterns of recovery
between people of different castes) the level of
accuracy of survey findings reduce. It should be noted
that the large sample size allows for more accurate
estimates, and that the margins of error are smaller
than in most surveys, in Nepal and beyond.
Perceptions and accurate reporting
The information provided throughout the report is
based on the reports of those interviewed. People may
have incentives to over- or under-report the level of
impact they experienced, and their perceptions or feel-
ings might not accurately reflect facts in some cases.
The data and findings should be read with this in mind.
But the use of both qualitative and quantitative re-
search has allowed for triangulation of findings, which
strengthens our confidence that they reflect reality.
4


Aid and Recovery in Post-Earthquake Nepal
1.3 Contextual changes since March 2016
A number of key contextual changes since IRM-2 was
conducted have shaped recovery.
The National Reconstruction Agency
At the time IRM-2 was being conducted, the National
Reconstruction Agency had just begun its work.6
The NRA is the lead government agency for all post-
earthquake reconstruction activities with a wide
mandate relating to coordination and facilitation of
reconstruction, recovery, and preparedness work.7
Formally established in late December 2015, political
wrangling over who would lead it mean that it only
just started to operate in practice at the time of IRM-2.
Since then, the NRA has become much more active
although it has also suffered from some technical and
political difficulties. Challenges faced include staffing
issues. The agency is reportedly facing difficulties in
attracting and retaining civil service staff. In December
2016, two months after IRM-3 fieldwork took place,
engineers deployed by the Ministry of Urban Devel-
opment Central Level Programme Implementation
Unit (MoUD CL-PIU) went on strike, citing poor
conditions. Additionally, the NRA has highlighted
shortages in technical staff and trained masons in
earthquake-affected districts. In response, the NRA
reached an agreement with the Nepal Army to mobi-
lize 200 army staff, including masons and carpenters
in Sindhuli, Okhaldhunga, and Ramechhap districts,
where there is an acute lack of skilled masons and tech-
nical manpower. The NRA has also started training
3,500 final year civil engineering students to assist in
reconstruction across the 14 most-affected districts.
Rural Housing Reconstruction Program
Delays in the establishment of the NRA meant that 1
little reconstruction work had been done at the time of d
IRM-2. Since then, the Government and donors have v
focused largely on housing reconstruction, in particu- (
larly through the Nepal Rural Housing Reconstruction r
Program (RHRP). t
c
The RHRP, which is supported by a multi-donor
fund, emphasizes owner-driven reconstruction, I
cash grants of NPR 300,000 are provided in three s
instalments to eligible beneficiaries to aid them in t
building earthquake-resistant houses. The June 2015 v
credit agreement between donors and the government t
requires the government to conduct a house-by-house t
damage assessment and eligibility survey, sign a f
participation agreement between eligible beneficiaries f
and the government, provide housing grants in three /
tranches through bank accounts, release subsequent a
tranches based on progress achieved in resilient i
construction and conduct comprehensive, multi-tier,
and hands-on training.
The new housing assessment, which began in some
districts in February 2016, around the time of IRM-2,
was conducted by the Central Bureau of Statistics
(CBS). The CBS initially deployed engineers to the 11
most affected districts, excluding districts categorized
by the government as being ‘hit with heavy losses’
or ‘hit’.
By September 2016, when IRM-3 was conducted, the
signing of beneficiary agreements and distribution of
the first installment of the reconstruction cash grant
was underway in 11 districts, including four visited in
the qualitative fieldwork and eight where the quanti-
tative survey was conducted. However, many people
had not yet received cash in hand. Since then, there
has been more progress with the RHRP program (see
Annex A). However, because these developments came
after fieldwork was conducted, they are not covered
in this report.
6 References for this section can be found in the IRM-3 qualitative
report.
7 The NRA is mandated to work closely with a number of other
government ministries. The Ministry of Federal Affairs and Local
Development, through its Central Level Programme Implemen-
tation Unit and District Level Programme Implementation Units,
holds primary responsibility for the disbursement of the housing
grant. Primary responsibility for technical standards and staffing
for housing reconstruction are the responsibility of the Ministry of
Urban Development (MoUD), through its CL-PIU and DL-PIUs,
as well as the Department of Urban Development and Building
Construction (DUDBC).
5


Introduction
1.4 Report structure
This report covers a number of areas:
• Progress in recovery. Chapter 2 considers chang-
es to and conditions in shelters, the progress
of reconstructing homes, progress in repairing
infrastructure and the status of service delivery,
the recovery of livelihoods, food provision and
needs, and trauma and vulnerability.
• Earthquake aid, critical needs, and housing re-
construction cash grants. Chapter 3 details the
nature of aid provided and how this has changed
over time, critical needs, experiences and levels
of satisfaction with assistance received and with
those providing it, and the coordination and
transparency of aid distribution.
• Coping strategies. Chapter 4 looks at how house-
holds have tried to cope with earthquake impacts
through financial behavior, migration, and other
means.
• Politics, social cohesion, and conflict. Chapter 5
reviews the extent to which the earthquake and
aid response have affected political party activ-
ities, roles, and levels of influence, changes in
people’s political preferences, and the impacts on
security, sources of conflict, and social cohesion.
Analysis of the differing impacts on different popu-
lation groups, differing patterns of recovery, and the
extent to which groups are vulnerable, is provided
throughout.
The report concludes with a summary of findings,
a discussion of implications for aid and recovery
efforts moving forward, and recommendations. These
conclusions and recommendations are not necessarily
those of the donors to IRM.
6


Photo: Anurag Devkota
2.1 Housing and shelter
Temporary shelters
The Nepal earthquakes had a devastating impact on
the housing stock in affected areas. In severely hit dis-
tricts, 79% of houses were completely destroyed and a
further 15% were badly damaged. Almost one year on
from the earthquakes, when IRM-2 was conducted,
80% of people in these districts were still living in
temporary shelters. Since then, government and donor
reconstruction programs have accelerated. How has
this affected the housing and shelter arrangements
of people?
Figure 2.1: Where people were/are living (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
7


Recovery
More people have moved back intopermanent
housing but the number of people in temporary
shelters is still high in severely hit districts.
There has been some progress in getting people back
into permanent housing. Over time, the number of
people living in temporary shelters has declined.
Figure 2.1 shows where people were living at the time
of the IRM-1 survey (June 2015), IRM-2 (February-
March 2016, and IRM-3 (September 2016). There have
been steady increases in the number of people living
in their own houses over time, and similar reductions
in the number of people in temporary shelters. As of
September 2016,71% of people in earthquake-affected
districts are in their own houses.
However, the picture is very different in severely
hit districts. Seventy-one percent of people in the
severely hit districts were still in temporary shelters
in September, one-and-a-half years from the disaster
(Figure 2.2). This figure has reduced since IRM-2
but, overall, there has been relatively little progress
in housing people in these districts. The number of
people still in temporary shelters is particularly high in
Sindhupalchowk (90%), Nuwakot (78%), Ramechhap
(73%), and Dhading (70%). The situation is somewhat
better in Gorkha, where over half of people are now
in their own homes. Amongst less affected districts,
Okhaldhunga has the highest proportion of people still
living in temporary shelters (25%).
Nine percent of people whose house was partially
destroyed and 52% whose house was completely
destroyed lived in self-constructed shelters as of
September. A small share of those whose house
was completely or partially destroyed now live in a
neighbor’s house or in shelter on other people’s land.
All people who reported minor or no damage to their
house from the earthquakes are now living in their
own houses.
Figure 2.2: Where people are living - by district impact (IRM-3, weighted)
Own house H Self-constructed shelter on own land
Neighbor’s house Self-constructed shelter on other people’s land
Peoplefrom marginalized groups are dispro-
portionately likely to still be living in tempo-
rary shelters.
More people in agricultural occupations live in
temporary shelters (48%) than is the case for other
livelihoods. Significant proportions of individuals with
a low income (47%) or no education (44%) continue
to live in shelters. Those with a disability are more
likely to be in shelters (38%) than those without (27%).
Higher proportions of Buddhists (46%) and Christians
(46%) still live in shelters.8
It should be noted that 17% of the population in affected areas is
Buddhist and only 1% are Christian.
Shelters are most commonly made of corru-
gated iron sheets (CGI) but in some districts
the proportion in other types of shelters is
comparatively high, most notably in Okhald-
hunga.
Among those who are living in shelters, the majority
are now in shelters fully made of CGI (62%). Over the
past six months, there has been a shift from people
living in shelters made partly out of wood, bamboo,
and CGI to those made of only CGI (Figure 2.3). Few
people are living in shelters that do not use CGI at all.
However, those in more remote areas are much less
likely to be in shelters made completely out of CGI.
In some districts, the proportion of those in shelters
made of other materials or in animal sheds is higher.
8


Aid and Recovery in Post-Earthquake Nepal
While only 4% of those in temporary shelters live
in shelters made of bamboo, the figure is higher in
Okhaldhunga (29%), Syangja (25%), Ramechhap
(16%), and Solukhumbu (14%). The proportion
of people in temporary shelters who are living in
cowsheds is relatively high in Okhaldhunga (10%) and
Lamjung (9%).
Figure 2.3: Share of people living in different types of shelters (IRM-2, IRM-3, weighted)
Wood/bamboo
â–  IRM-2 â–  IRM-3
Movements to and from temporary shelters
Movements to and from shelters have been
significant between all research rounds.
While many people have left their temporary
shelters to move back into their houses, some
have also since moved out again, returning to
temporary shelters or moving into the homes
of others.
The pace of people moving to and from shelters has
been similar between IRM-2 and IRM-3 compared to
the first year after the earthquakes. One-quarter of
those who were in self-constructed temporary shelters
at the time of IRM-1 (June 2015) had moved back
into their own house by the time of IRM-2 (February-
March 2016) and the figure is 36% for those who were
in community shelters during IRM-1. At the same
time, large numbers of people moved out of their own
house between IRM-1 and IRM-2 often returning to
temporary shelters.9
By IRM-3 (September 2016), 24% of those who were
living in shelters on their own land at the time of
IRM-2 (February-March 2016) were able to move
to their own houses with the figure 18% for those
living in shelters on others’ land at the time of IRM-
2. Twenty-one percent of those who were living in
their own homes in IRM-3 had been living in self-
constructed shelters in IRM-2. As in the first year after
the earthquakes, however, there was also movement of
some people who were in houses during IRM-2 back
into temporary shelters by the time of IRM-3. Eleven
percent of those who were in their own house at the
time of IRM-2 were living in temporary shelters by
the time of IRM-3. Twelve percent of those who were
living in shelters on their own land at the time of IRM-
3, 9% of those who were in shelters on others’ land,
and one-third of those in shelters on public land had
been in their own house at the time of IRM-2.
The proportion of people who were in shel-
ters who have moved home varied between
districts.
Overall, 22% of people who had been in temporary
shelters at the time of IRM-1, in the weeks after the
earthquakes, were in their own house by IRM-3. In
9 Six percent of those who had been in their own house at the time
of IRM-1 were living in temporary shelters by the time of IRM-2.
Half of the people living in a friend’s house at the time of IRM-
2 had been living in their own house in IRM-1. Four percent of
people who were living in temporary shelters on their own land
had been in their own house at the time of IRM-1, and the figure
is 3% for those in shelters on others’ land. Twenty-seven percent
of people who were renting at the time of IRM-2 had been in their
own house during IRM-1.
9


Recovery
Solukhumbu, Okhaldhunga, and Gorkha, over 40%
of those in shelters in IRM-1 reported that they had
moved back to their own house by IRM-3 (Table 2.1).
Solukhumbu has the highest rate of any district of
people having fully repaired/rebuilt their houses or
built a new one (31%) while Okhladhunga also ranks
high (20%). However, the gap in both districts between
the number of people who have moved home and
those who have repaired or rebuilt suggest that, as
elsewhere, people are moving into potentially unsafe
houses. In contrast, only 6% of those who were in
temporary shelters during IRM-1 have moved back to
their own home in Syangja, the least affected district in
the sample, and the figure is also low for Kathmandu,
Sindhupalchowk, and Lamjung.
Many of those who returned to their houses
were moving into unsafe buildings, often after
no or only minor repairs, or to at-risk land.
For this reason, some have since moved out
again.
While the movement of many to their own home at
first looks promising, many maybe moving into unsafe
housing. Table 2.2 shows that while most people who
have done nothing to repair or rebuild their house
remain in temporary shelter, 17% have moved home,
suggesting the structure they are moving in to may not
be safe. Further, almost two-thirds of those who were
in temporary shelters who have started (re)building,
but whose house is not yet finished, have moved
home. And almost one-quarter of those who were in
temporary shelters who have started rebuilding, but
who acknowledge their house is not yet livable have,
despite this, moved into their house.
Table 2.1: Share of people who were in
shelter (IRM-1) to their own house (IRM-3) -
by district impact and district (IRM-1, IRM-2,
IRM-3 household panel, unweighted)
Moved from shelter to house
Severely hit 22%
Dhading 21%
Gorkha 42%
Nuwakot 14%
Ramechhap 23%
Sindhupalchowk 12%
Crisis hit 29%
Bhaktapur 18%
Kathmandu 11%
Okhaldhunga 48%
Hit with heavy losses 19%
Lamjung 13%
Solukhumbu 44%
Hit 6%
Syangja 6%
AII districts 22%
Table 2.2: Share of people who moved from shelter (IRM-1) to their own house (IRM-3) - by what people
have done to their house (IRM-1, IRM-2, IRM-3 household panel, unweighted)
Moved from shelter to house Have not moved from shelter to house
Have done nothing to rebuild it/build new house 17% 83%
I have fully repaired/rebuilt my house and I live in it now 43% 57%
I have built a new house 62% 38%10
I have partly rebuilt/built a new house. It is not yet finished but I live in it 62% 38%
I have started to rebuild/build a new house but it is not yet livable 24% 62%
This explains why some have since moved out again;
people moved from their own house to other accom-
modation options so that they could rebuild their
house, or simply to stay somewhere safer after realiz-
ing that their house was not safe.
The qualitative research also found that increasing
numbers of people were moving back into unsafe
houses. In 20 of 36 wards visited, researchers came
across a significant number of households moving
back into damaged houses after only minor repairs
and in 10 wards several households were observed
to be moving back into houses without any repairs.
Yet, as findings from the survey also highlight, some
households have since moved out again, returning
10 This suggests some people split their time between living in a
shelter and in their own home. For example, they may sleep in
the shelter but cook in their house.
10


Aid and Recovery in Post-Earthquake Nepal
to temporary shelters or moving elsewhere, after
realizing that their houses were not safe enough or
could not withstand bad weather.11 Some even saw
their houses collapse during the monsoon rains.
“Some of those houses that only had minor cracks
from the earthquake collapsed during the monsoon.
Recently there was one incident [...] where a partially
damaged house collapsed, badly injuring the single
woman living there,” said a resident of Katunje VDC
in Okhaldhunga. Other highlighted their fears of living
in houses after doing only minor repairs. “Repaired
houses seem good from the outside if painted but they
are risky from the inside. One can see cracks from the
inside and we feel fear living in these houses,” said a
55-year old woman in Dudhkunda municipality of
Solukhumbu. Similarly, a man in Syangja exclaimed,
“We need money as promised by the government so
that we can construct a new house. Living in damaged
houses never gives us peace, we are always in fear.”
Further, the qualitative research highlighted that some
of those displaced from their land due to landslide
risks were also moving back or planning to move back
soon, despite the danger. They were willing to return to
at-risk land due to ongoing uncertainty over long-term
resettlement solutions and discomfort in temporary
settlements or tensions with local communities in
their new settlements.
Preparedness for adverse weather
Of the people in temporary shelters, those
in remote regions, rural areas, and severely
or crisis hit districts were less prepared for
the monsoon. In particular, marginalized
groups, including those with disability, were
less likely to be prepared for adverse weather.
The proportion of people who were able to completely
fix their accommodation to withstand the weather
(winter in IRM-2 and monsoon in IRM-3) increased
from IRM-2 to IRM-3 - from 3% to 6%. However,
the proportion of people who were not able to make
repairs at all also rose - from 6% to 17%. The most
common reason why shelters were deemed insufficient
in IRM-3 was that they had leaky roofs or walls, while
many also said their shelter was too cold for living.
More people in more remote regions, rural areas, and
severely and crisis hit districts were less prepared for
monsoon. Among the districts, relatively more people
(more than 20%) in Sindhupalchowk, Okhaldhunga,
Lamjung, Gorkha, and Dhading reported either in-
sufficient or no repairs for the monsoon. Only 34%
of people in Sindhupalchowk said they had been able
Illnesses in temporary shelters
Many got sick during the monsoon months
due to issues with shelter.
Twelve percent of people interviewed in the survey said
that they, or someone in their family, got sick during
to make sufficient repairs for the monsoon, with 48%
saying they had made no repairs at all.
Marginalized groups were less prepared for the mon-
soon than others. Two-thirds of low caste people and
Janajatis said they had been unable to make sufficient
repairs, or had made one, compared to 75% of high
caste respondents. Those with disabilities (38%) were
more likely to be unprepared for the monsoon com-
pared with those without (27%).
Those whose houses were fully destroyed,
those living in severely hit districts, and those
with low incomes have been consistently
unprepared for adverse weather.
Those unprepared for the winter and the monsoon
in IRM-2 and IRM-3 were those whose houses had
been completely destroyed (96% of those unprepared
in both rounds), those living in severely hit districts
(89% of those unprepared in both rounds), and those
who have a low income (69% of those unprepared in
both rounds).12
the monsoon because of problems with shelter. This
figure is much higher (23%) in severely hit districts.
Incidence of illnesses due to shelter issues during the
monsoon was particularly high in Nuwakot (45%). Just
over one-fifth of respondents report a shelter-related
11 This was already observed at the time of IRM-2. See, The Asia
Foundation and Democracy Resource Center Nepal (2016). Aid
and Recovery in Post-Earthquake Nepal: Independent Impacts
and Recovery Monitoring Phase 2 - Qualitative Field Monitoring
(February and March 2016). Kathmandu and Bangkok: The Asia
Foundation.
12 Analyzing the household panel data from IRM-2 and IRM-3 allows
us to assess the section of the population who were not ready for
both the winter and the monsoon. To assess this, respondents in
the household panel dataset of the last two rounds were labelled
as “not ready” if they chose in both waves either “they were not
able to repair at all” or “even if they repaired, it was not sufficient.”
11


Recovery
illness in the family in Gorkha (20%), Ramechhap
(24%), and Sindhupalchowk (21%).
Those living in rural areas (15%) were more likely
than people in urban areas (6%) to have someone in
the family who fell ill during the monsoon due to their
shelter. Those in more remote (16%) or remote (13%)
areas were more likely to have someone in the family
fall ill during the monsoon than in less remote areas
(9%). People belonging to marginalized groups—those
with lower incomes (19%), women (14%), and the
disabled (23%)—were more likely to report a shelter-
related illness in the family during the monsoon.
Those who said that the earthquake destroyed their
house completely (18%) were more likely than those
reporting less damage to say someone got sick during
the monsoon disease. Further, those who were
unable to make any repairs to their house in order to
get it ready for the 2016 monsoon (34%) are much
more likely to report an illness in the family than
people who were able to get some level of repairs
done. People living in communal or self-constructed
shelters were more likely than those living in houses
to report someone in their family getting sick during
the monsoon due to their accommodation. A majority
of those living in a community shelter (56%) reported
someone getting sick during the monsoon. Among
those who lived in a self-constructed shelter, those
who built it on public land (46%) were more likely
than those who built it on others’ land (28%) or on
their own land (25%) to say there was an illness. In
contrast, fewer people who lived in a house, whether a
friend’s (22%), a neighbor’s (17%), or their own (6%),
said that someone in their family got sick.
Common illnesses were colds, fever, and stomach issues.
Of the 12% who reported an illness in the family, fever
(54%) and recurrent colds (34%) were the most com-
mon ones. Far fewer mention prolonged colds (12%),
swollen feet (9%), diarrhea/dysentery/cholera (8%),
pneumonia (5%), asthma (5%), or skin rashes (3%).
In the qualitative research, various health problems
associated with living in shelters were also observed.
Respiratory diseases, such as asthma among the
elderly and pneumonia among children, as well
as diarrhea and dysentery due to poor sanitary
conditions, continued to be observed. Some thought
that not enough attention was paid to the suffering
and illnesses of those in shelters. As a local activist
and entrepreneur in Dudhkunda municipality in
Solukhumbu said: “Those people who can afford to
are rebuilding their houses, but the majority of the
families whose houses were destroyed are living
in temporary shelters. The weather conditions are
very harsh, we have had many cases of asthma and
pneumonia. People have died of this in our region,
but it is not noticed.”
Cases of malnutrition among children in
temporary shelters were reported to have in-
creased in Okhaldhunga and parts of Gorkha
and Sindhupalchowk.
The qualitative research suggests that the increase was
due to changing food habits in temporary shelters and
lowered harvests after the earthquakes. Sovita Dahal,
an Assistant Nurse and Midwife from Prapcha VDC in
Okhaldhunga, explained: “Malnutrition has slightly
increased after the earthquake [...]. Across the VDC, two
children are suffering from hard malnutrition and eight
are suffering from mild malnutrition. There was only
one such case before the earthquake. After the earth-
quake, parents are not able to follow the feeding sched-
ule for their children. As a result, children are not getting
enough nutrition.” Other respondents, too, pointed out
that the earthquake had changed people’s eating habits.
Reconstruction of houses
As of September 2016, progress in the recon-
struction of homes had been slow. Of those
whose house was impacted, most people had
done nothing to repair or rebuild.
Seventy-two percent of the respondents whose
house was impacted say that as of IRM-3 they have
done nothing in terms of repairing or building new
houses (Table 2.3). This response was much higher
in severely hit districts, where 80% report not having
done anything to repair their damaged house or to
build a new house. Ten percent of people in severely
hit districts whose house was impacted have either
repaired it or built a new one. The share of people
who have done nothing to repair or rebuild their
house is higher in remote areas (76%) and in more
remote regions (72%) compared to less remote regions
(66%). Among severely hit districts, Gorkha has the
lowest share of people (73%) who report no progress
in repairing their existing house or building a new one
but this figure is still very high.
Lower caste and low income groups are less
likely to have repaired or rebuilt their houses.
Of those whose house was impacted, 84% of people
with a disability and 74% with a low income say that
they have done nothing to repair or rebuild (Table 2.4).
Disaggregating by caste, 72% of Janajatis and 82% of
low caste people say they have not taken any actions
to repair or rebuild.
12


Aid and Recovery in Post-Earthquake Nepal
Table 2.3: Actions to repair or rebuild houses amongst those whose house was impacted -
by district impact, district, rural/urban and remoteness (IRM-3, weighted)
Have done nothing to rebuild it/ build new house 1 have fully repaired/ rebuilt my house and 1 live in it now I have built a new house I have partly rebuilt/ built a new house. It is not yet finished but I live in it I have started to rebuild/build a new house but it is not yet livable Refused/ don’t know
Severely hit 80% 8% 2% 7% 4% 0%
Dhading 78% 17% 1% 3% 1% 0%
Gorkha 73% 4% 4% 15% 4% 0%
Nuwakot 86% 10% 1% 4% 0% 0%
Ramechhap 82% 5% 2% 11% 1% 0%
Sindhupalchowk 80% 4% 2% 3% 11% 0%
Crisis hit 67% 18% 2% 8% 3% 1%
Bhaktapur 69% 12% 3% 8% 8% 0%
Kathmandu 67% 20% 1% 8% 2% 2%
Okhaldhunga 68% 15% 5% 11% 2% 0%
Hit with heavy losses 65% 28% 3% 2% 2% 0%
Lamjung 63% 29% 1% 5% 2% 0%
Solukhumbu 67% 27% 4% 0% 2% 0%
Hit 53% 43% 1% 2% 1% 1%
Syangja 53% 43% 1% 2% 1% 1%
All districts 72% 15% 2% 7% 3% 1%
Rural areas 72% 15% 2% 8% 3% 0%
Urban areas 72% 18% 1% 5% 2% 2%
Less remote 66% 22% 1% 6% 3% 2%
Remote 76% 13% 2% 6% 3% 0%
More remote 72% 11% 3% 11% 3% 0%
Table 2.4: Actions to repair or rebuild houses amongst those whose house was impacted -
by gender, caste, income and disability (IRM-3, weighted)
Have done nothing to rebuild it/ build new house I have fully repaired/ rebuilt my house and I live in it now I have built a new house I have partly rebuilt/ built a new house. It is not yet finished but I live in it I have started to rebuild/build a new house but it is not yet livable Refused/ don’t know
Female 73% 16% 1% 7% 3% 0%
Male 71% 15% 2% 7% 3% 1%
High caste 71% 16% 1% 8% 2% 1%
Janajati 72% 15% 2% 6% 4% 0%
Low caste 82% 9% 2% 7% 1% 0%
Low income 74% 12% 2% 7% 4% 0%
Medium income 74% 16% 1% 6% 2% 0%
High income 64% 19% 2% 9% 3% 3%
No disability 72% 16% 2% 7% 3% 1%
Disability 84% 10% 1% 2% 2% 0%
13


Recovery
A lack of money was the primary factor that
prevented people from starting to rebuild
their houses. Other commonly cited reasons
were people waiting to receive the recon-
struction cash grant or a lack of knowledge
on approved building designs.
The primary reason why many people did not start
rebuilding was a lack of money.13 Eighty-nine percent of
people who had not yet rebuilt cited not having enough
money as the reason why (Table 2.5). Similarly, 66% of
people were waiting for the government distribution
of cash grants, with percentages citing this higher in
severely hit districts.14
Thirteen percent said they had not yet rebuilt because
they were unsure what types of houses are allowed by
the government and 7% because they had not been giv-
en instructions on how to build a safe house. Receipt of
subsequent tranches of government cash for rebuild-
ing is dependent on houses being earthquake-proof
and following government-approved guidelines. A
previous study and the qualitative research has shown
that there was little knowledge of what the rules are
and that this, combined with limited technical assis-
tance, has hampered rebuilding efforts.15 Six percent of
people who have not rebuilt say that a lack of labor is
a problem. This is particularly a problem in Nuwakot,
where 34% say it has prevented them from rebuilding.
Unsurprisingly, the poor are more likely to say that a
lack of money has prevented them rebuilding (93%).
The poor are also more likely to say they are waiting
for government cash grants and that the price of con-
struction materials is too high.
Table 2.5: Reasons for stopping repairing or not building a house - by district impact,
district, remoteness, rural/urban and income (IRM-3, weighted)
Did not have enough money Still waiting for government cash grant Unsure what types of houses are allowed by the government Still waiting for instructions on how to build safe house Still waiting for geological assessment No labor to rebuild Prices of construction materials too high No family members around to help Do not have land related papers Refused/don’t know
Severely hit 92% 84% 19% 10% 4% 8% 15% 3% 0% 0%
Dhading 94% 73% 20% 18% 9% 0% 12% 3% 0% 0%
Gorkha 86% 83% 15% 4% 0% 2% 7% 4% 0% 0%
Nuwakot 100% 93% 18% 6% 5% 34% 43% 1% 0% 0%
Ramechhap 85% 86% 17% 6% 0% 3% 3% 3% 0% 0%
Sindhupalchowk 95% 85% 24% 15% 5% 5% 11% 3% 1% 0%
Crisis hit 87% 51% 5% 2% 0% 1% 15% 0% 1% 2%
Bhaktapur 97% 59% 21% 7% 1% 1% 26% 1% 1% 0%
Kathmandu 84% 45% 0% 1% 0% 0% 11% 0% 1% 2%
Okhaldhunga 94% 76% 10% 5% 0% 6% 26% 3% 0% 1%
Hit with heavy losses 76% 45% 14% 9% 3% 11% 32% 6% 0% 1%
Lamjung 65% 54% 16% 18% 7% 10% 26% 3% 0% 2%
Solukhumbu 86% 37% 12% 1% 0% 12% 36% 9% 1% 0%
Hit 91% 4% 2% 2% 0% 4% 2% 1% 0% 2%
Syangja 91% 4% 2% 2% 0% 4% 2% 1% 0% 2%
All districts 89% 66% 13% 7% 2% 6% 16% 2% 0% 1%
Less remote 88% 50% 8% 4% 1% 2% 16% 0% 1% 2%
Remote 90% 73% 14% 8% 3% 8% 16% 3% 0% 1%
More remote 89% 77% 17% 8% 3% 3% 15% 3% 1% 0%
15 See The Asia Foundation and Democracy Resource Center Nepal
(2016). Nepal Government Distribution of Reconstruction Grants
for Private Homes: IRM - Thematic Study (November 2016).
Kathmandu and Bangkok: The Asia Foundation.
13 People could give multiple reasons, hence percentages do not
add up to 100%.
14 At the time of research, cash grants were being distributed in
only the 11 most affected districts. The fact that very few people
cited this as a reason in Syangja, the least affected district in the
sample, suggests that people there may have had little expectation
that government cash grants will reach them.
14


Aid and Recovery in Post-Earthquake Nepal
Did not have enough money Still waiting for government cash grant Unsure what types of houses are allowed by the government Still waiting for instructions on how to build safe house Still waiting for geological assessment No labor to rebuild Prices of construction materials too high No family members around to help Do not have land related papers Refused/don’t know
Rural area 91% 70% 14% 8% 3% 7% 16% 2% 1% 1%
Urban area 81% 52% 5% 3% 0% 1% 17% 0% 0% 2%
Low income 93% 78% 13% 6% 2% 10% 20% 3% 1% 0%
Medium income 88% 63% 14% 8% 3% 3% 14% 1% 0% 1%
High income 82% 51% 11% 8% 3% 3% 8% 1% 0% 1%
Costs for the construction of houses were in-
creasing significantly.
Sixteen percent of those surveyed said the high price
of construction materials was a reason why they had
not rebuilt. When asked if there had been changes
in the costs of construction materials since the end
of last winter, 92% said that the cost of construction
labor was higher than before, 85% mentioned that
construction material had become more expensive,
and 87% mentioned that CGI sheets were now costlier
(Table 2.6).
Table 2.6: Cost of construction materials (IRM-3, weighted)
Much higher Slightly higher Same Slightly less Much less Refused Don’t know
Cement 43% 34% 2% 1% 1% 1% 18%
Iron rod 45% 32% 1% 2% 1% 2% 17%
Stone/bricks 42% 38% 3% 1% 1% 2% 14%
Wood/Timber 41% 40% 5% 0% 0% 3% 11%
Nails 34% 51% 4% 0% 0% 2% 9%
CGI 42% 45% 3% 0% 0% 2% 7%
Tiles 33% 28% 3% 0% 1% 5% 29%
Construction labor 53% 39% 1% 0% 0% 2% 5%
Construction materials 38% 47% 1% 0% 0% 2% 12%
In the qualitative research, too, people across the
districts visited pointed out that construction costs had
increased compared to pre-earthquake times. In many
remote places, they were drastically higher. Materials
that are often not locally available in rural Nepal,
such as cement, bricks, iron rods, corrugated iron, or
sand, were particularly expensive. The high demand
for these materials—they have to be used to rebuild
houses according to the approved building codes for
earthquake-resilient buildings16—also increased costs.
In many areas, even locally available materials such
as wood or bamboo had become more expensive due
to high demand and restricted access to community
forests.17
High transportation costs for construction
materials was one of the main reasons why
construction costs had increased.
In many areas, people complained that construction
materials were not available locally and had to be
transported from the district headquarters or other
hubs. In remote places, especially those without
roads, costs were particularly high. This often meant
that those trying to rebuild had to spend much of
their money on transportation. For example, each
truck or tractor transporting sand from Manthali,
Ramechhap’s district headquarters, to Bamti Bhandar
VDC cost NPR 10,000. The first installment of the
cash grant (NPR 50,000) was therefore insufficient to
construct even one pillar, complained locals.
16 These guidelines have to be followed if homeowners want to
receive further installments of the reconstruction cash grant for
private houses.
17 See Case Study 7.2, ‘Community forests and the use of local
resources for reconstruction’, in the IRM-3 qualitative report.
15


Recovery
Heavy rains meant that the transportation of goods
became even more difficult and expensive during the
monsoon. Trucks and other larger transporters could
no longer drive on roads rendered inaccessible by mud
or landslides. The quality of roads determined whether
transportation was more difficult, and hence more
expensive, during the monsoon. In Solukhumbu, for
example, rising transportation costs for construction
materials and other goods affected all wards visited,
including remote wards of the district headquarters,
as most roads are temporary and regularly damaged
during the monsoon. With some of the earthquake-
affected villages more than a three-day walk from
the district headquarters, transportation of aid and
construction materials has been one of the major
reconstruction-related difficulties in the district (see
Case Study 2.1).
Other common reasonsfor rising construction
costs were high wages for laborers and water
shortages.
High wages for laborers also raised overall costs for
house reconstruction. Due to the high demand for
construction laborers, their wages had increased
significantly (see Chapter 2.3). Water shortages are also
raising costs. Water is needed in large amounts for the
construction of cement houses and some households
have had to buy water for the construction of their
houses. For example, in Ramechhap municipality one
resident had to spend NPR 65,000 to buy water to
construct his house. Given that water shortages were
common across wards visited, it is likely that many
more will have to pay for water for the construction
of their houses.
Case Study 2.1: High transportation costs during
the monsoon affect reconstruction
The house of Pranay Rai from Nele VDC, which
was made from mud, stone, and wood, was
damaged during the second major earthquake
of 12 May 2015. He decided to build a semi-
concrete, earthquake-resistant house before
the 2016 monsoon. “It looked like I would have
to wait for years to receive anything from the
government, and living in a half-damaged house
was risky and also hampering my [electronics]
business,” he said. “So I arranged for some loans
from local saving organizations, and some from
my friends and family, to add to my savings and
started building the house in June.”
The timing for the construction of his new house
meant that transportation costs for construction
materials were unusually high, having increased
drastically during the monsoon. Nele VDC,
although an important market center about
18 kilometers east of the district headquarters,
is only connected by a dirt road, which is
frequently affected during the rainy season.
Only tractors can occasionally drive on the road
during the monsoon.
“We transport cement from Okhaldhunga
Bazaar [about 65 kilometers from Solukhumbu’s
district headquarters] at two Rupees per kg
during the dry season when the transportation
is normal and uninterrupted, but now we pay
seven Rupees per kg,” Rai said. This means
an additional NPR 100-250 per 50 kilogram
sack of cement - an increase of 150 percent.
The transportation cost of other materials
like tin sheets, iron rods, and metals were
also disproportionately high in Nele during
the monsoon. “I will have spent over one lakh
extra only in transportation costs by the time I
complete the house,” Rai complained. “What
use is the two lakhs that the government is
throwing at us?”
Rai can afford the higher transportation costs
due to a regular income from his electronics
business. But many other households in the
VDC have fewer resources. Tilak BK from Nele
VDC said, “Everything has become extremely
expensive - from wages for masons and
construction workers to wood and stones, and
transportation. I have to look after a family of
five and, I don’t have a regular source of income
as I depend on daily wage labor for half of
the year. If the government does not provide
us assistance, I will have to bear the burden
of the loan for years.” He said this despite
having received two lakhs in cash assistance
from an individual donor who supported the
rebuilding of many of the damaged houses in
Nele. “We got two lakhs but ended up spending
three more lakhs which we had to borrow
from moneylenders and savings groups,” he
explained.
16


Aid and Recovery in Post-Earthquake Nepal
Photo: Chiran Manandhar
2.2 Infrastructure and service delivery
Access to services has improved since the
early months after the earthquakes.
Almost everyone surveyed in IRM-3 said that elec-
tricity, drinking water, access to a medical facility,
schools, and motorable roads was provided by VDCs
and municipalities (Figure 2.4). There were particular
improvements in the provision of drinking water and
medical facilities.
Figure 2.4: Share saying they have the services provided by VDC/municipality
(IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
17


Recovery
For most services, a large share say the quality has
improved since IRM-2 - Figure 2.5. Over 50% say
that medical facilities and schools have improved and
almost half note the same for motorable roads. Around
one-third say that drinking water has improved. For
electricity, 27% say it has improved, but 19% say it
has worsened.
The qualitative research highlights that reconstruction
of community infrastructure, roads, water and sanita-
tion facilities, health care, and improvements of school
infrastructure remained frequently identified and ur-
gent needs. Water shortages and the lack of irrigation
remained particularly common problems despite good
monsoon rains. Despite improvements in services and
infrastructure, needs in these areas remain.
Figure 2.5: Changes in quality of services (IRM-1, IRM-2, IRM-3, weighted)18
IRM-3

IB S 5^ IRM-2
IRM-1
bJO s IRM-3
£ (3 "cb IRM-2
•C £
Q IRM-1
IRM-3
IRM-1
£
CD
w
IRM-3
O 0 IRM-2
O £ IRM-1
IRM-3
o
o
43
Q
C/D
3%
44%
35%
18%
IRM-2
IRM-2
IRM-1
12% 36% 52%
6% 20% 72%
3% 9% 55% 10% 22%
18% 43% 39%
15% 27% 57%
5% 14% 54% 6% 21%
13% 42% 44%
15% 31% 54%
4% 9% 37% 34% 16%
15% 37% 47%
9% 22% 69%
3% 39% 37% 20%
22% 33% 45%
39% 34% 26%
1 1 1 1 1 1
20%
40%
60%
80%
Worsened a lot
Somewhat worsened
| Nothing has changed I A lot better
Somewhat better
0%
J
100%
Most were satisfied with the services they
were getting but overall, satisfaction with
services provided was decreasing.
Satisfaction with all five services has dropped since
IRM-1, though people were more likely to be satisfied
than dissatisfied with each service (Figure 2.6). There
was not much change in satisfaction levels between
IRM-2 and IRM-3. For instance, 89% were satisfied
with electricity at home in IRM-1 compared to 60% in
IRM-2 and 63% in IRM-3. Schools are the exception.
Though satisfaction with schools dropped in IRM-2,
it rose to 90% in IRM-3, quite close to satisfaction in
IRM-1 (93%).
18 In IRM-3 the response options ‘somewhat better’ and ‘a lot better’
were added to the question.
18


Aid and Recovery in Post-Earthquake Nepal
Figure 2.6: Satisfaction with public services (IRM-1, IRM-2, IRM-3, weighted)
S—< —‘
O
+-•
O
£
o
o
43
u
co
gb
cd
WO
•g £
a £
•C £
Q
b

IRM-3
IRM-2
IRM-1
IRM-3
IRM-2
IRM-1
IRM-3
IRM-2
IRM-1
IRM-3
IRM-2
IRM-1
IRM-3
IRM-2
IRM-1
45% 35% 10% 7% 3%
50% 30% 5% 11% 4%
70% 21% 4% 3%2%
51% 39% 7% 2%1%

53% 32% 4% 8% 3%
68% 25% 3%2%2%
42% 38% 13% 6% 1%
| 45% 36% 5% 10% 4%
69% 25% 4%1%1%
20% 47% 17% 12% 4%
I 27% 34% 6% 15% 18%
I 52% 33% 8% 3% 4%
16% 47% 19% 14% 4%
32% 28% 5% 12% 24%
I 60% 29% 5% 4% 2%

% 20% 40% 60% 80% 100%
2
Very satisfied H Neither satisfied nor dissatisfied Very dissatisfied
Somewhat satisfied H Somewhat dissatisfied
2.3 Livelihoods
Recovery of livelihoods
Different livelihoods continued to recover,
with recovery more widespread in mid-2016
compared to early 2016.
The predominant income sources in districts affected
by the earthquakes are farming and business.19 * * * 19 In
IRM-2, those who worked in business (72%) or who
were daily wage laborers (59%) were the most likely
to state that their income was negatively impacted by
the earthquakes. In the severely hit districts, the most
widely impacted occupation was farming, with 75%
of those who farmed their own land saying that their
income had been negative impacted.20 Around half
of those affected said that their income source had
improved in the first quarter of 2016, with the pro-
portion of people reporting recovery varying between
income sources.
The IRM-3 data show that recovery was more wide-
spread in mid-2016. For every source of income,
19 Across all districts, farming was a significant source of income
before the earthquakes for 58% of people and business for 37%.
Farming is particularly important in the severely hit districts,
where 96% report it as a major source of income and in more
remote areas (97%). Business ownership is much more common in
the crisis hit districts, which include Kathmandu and Bhaktapur,
and in less remote regions. Other common sources of income are
livestock farming (21%, 46% in severely hit districts), daily wage
work (17%, again more common in severely hit districts) and
salary work for private companies (15%, more common in the
urban crisis hit districts). Data are from the IRM-3 survey. People
could report more than one source of income.
20 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 2: February-March 2016. Quantitative Survey.
Kathmandu and Bangkok: The Asia Foundation, p. 25, pp. 10-12.
19


Recovery
a much larger proportion of people who said their
income source had been negatively impacted by the
disaster said they have seen (some) recovery in the
third quarter of 2016 compared to IRM-2 (Figure 2.7).
For example, while 53% of those who generate income
from farming their own land whose income was dam-
aged by the earthquakes said they had not seen any
recovery in the first quarter of 2016, only 15% report
the same for the last three months.
Those in severely hit districts were at least as likely
to see recovery of their income source as those
in other districts. Eighty-four percent of those in
severely hit districts whose farming was affected
said they have seen recent improvements. Amongst
severely hit districts, farmers tilling their own land
in Ramechhap and Sindhupalchowk less commonly
reported improvement in their income (78% and 73%,
respectively).
Businesses, too, were recovering. More than 90% of
business owners who were affected by the disaster
reported improvement in every district, with the
exception of Kathmandu (79%) and Lamjung (67%).
Recovery for business owners varied systematically
across levels of remoteness. Those in more remote
regions were doing much better (95% had seen
recovery in the three months prior to September 2016)
compared to those in remote areas (89%) and less
remote regions (79%).
Figure 2.7: Share of people within each income source whose income from that source
has improved in the last three months - by source of income (IRM-2, IRM-3, weighted)
â–  IRM-2 â–  IRM-3
Small business owners who lost their business lo-
cations as well as goods during the earthquakes,
however, were continuing to struggle as they had not
been given any compensation for their losses. This par-
ticularly affected poorer and women business owners
who had established small shops or restaurants with
loans or micro credit in rented places but who lacked
the resources to reestablish their businesses without
additional support to rebuy goods and reopen in new
locations. Business owners continued to highlight that
damages to businesses should have been assessed and
compensation provided by the government.
Tourism businesses were beginning to see notable
improvements for the first time since the earthquakes.
Tourism was picking up again with increased numbers
of visitors and good bookings for the upcoming season.
However, full recovery was not expected any time soon
by those working in the sector.
Laborers working in reconstruction saw some of the
biggest increases in wages as demand was rising with
the beginning of wide scale reconstruction. As seen
in Figure 2.7, daily wage labor and those working in
private companies saw significant improvements in
mid-2016. Skilled laborers such as carpenters and
masons, as well as some unskilled laborers, were
observed to be benefitting from more opportunities
and increasing wages after the earthquakes (Case
Study 2.2). In rural Nepal, wage labor has traditionally
been only a complementary economic activity for
famers. With fewer households relying primarily on
agriculture, labor at home and abroad has increased
in importance. As such, the fact that labor was only
temporary interrupted and that demand for, and
incomes from, wage labor are rising, especially in the
construction sector, are encouraging signs for the
recovery of earthquake-affected families.
20


Aid and Recovery in Post-Earthquake Nepal
Case Study 2.2: Masonry continues to gain in Solukhumbu
Laxman Basnet from Kerung VDC in Solukhum-
bu previously had to go to the high altitude vil-
lages of Khumjung and Namchhe for about four
months each year to look for work as a semi-
skilled mason. “There are too many masons in
our village—almost every other household has
one—and it was difficult to get regular work in
the village before the earthquake/’ Basnet said.
“But after the earthquake, there is so much
demand, I haven’t had a single day off.” Basnet
explained that one head mason would at most
build two houses per winter but now they were
building up to four. Basnet also said that there
had been a gradual increase in wages since the
earthquake. “We used to work for about NPR
800 per day but now people are earning up
to NPR 1,250 per day.” Masons from Kerung
are also working in other VDCs, some as far as
Gorkha, where the daily wages are reported to
be even higher.
Asked if he was familiar with the government-
provided guidelines for earthquake-resistant
houses, Basnet said that he had only heard
about them on the radio but had never seen a
model house. “I think it would be really useful
for us if the government built a model house in
the village and gave us some trainings. We learn
easily through experience and I am sure trained
local masons could help in the reconstruction
process.” But he was convinced that the houses
that are being built locally after the earthquakes
are much safer and stronger than before. “We
are now using extra safety with wooden bands
in the joints and the roof, and people also do
not want to build very tall.”
Challenges for farmers
Many farmers mere struggling and in need of
support despite disruptions to farming being
mostly restricted to the early weeks after the
earthquakes.
Farming was only temporarily disrupted in most
places and farmers’ ability to cope in the aftermath of
the earthquakes was enhanced by the fact that most
also relied on other sources of income such as small
businesses, daily wage labor, or migration. Exceptions
were farming households that lost members during
the earthquakes due to the lack of manpower and
psychological impacts. Other factors directly related
to the earthquakes that continued to affect farming
in mid-2016 were damage to agricultural land and
landslides risks, displacement and long commutes
from shelters to the fields, a lack of space to store
harvests, the construction of temporary shelters
on cultivable land, the loss of animals and reduced
availability of manure, and the option to earn higher
wages from construction work. The drying up of water
sources and damages to, or the absence of, irrigation
were also commonly cited but could not always be
directly attributed to the earthquakes.
Over the longer term it has become clear that farmers
are facing significant difficulties, many of them not
earthquake-related, that may prevent full recovery.
As reported in IRM-2, general hardships faced by
farmers in rural Nepal were exacerbated by the earth-
quakes.21 Pre-existing conditions of poverty and other
factors such as water shortages and a lack of irrigation,
have become more significant since the earthquakes,
making it even harder for farmers to overcome the
consequences of the quakes and other obstacles. It
is therefore unsurprising that farmers in the wards
visited in the qualitative research said that yields have
decreased, sometimes up to 30 percent, but that they
were often unable to distinguish whether this was
because of the impact of the earthquakes or due to
other unrelated difficulties.
Earthquake impacts continued to affect live-
stockfarming and reduced the availability
of manure.
Most households involved in agriculture are also
engaged in animal husbandry, which continued to
experience the impacts of the earthquakes due to
losses of livestock, collapsing of sheds, limited space in
temporary settlements, and limited water and fodder
for the remaining animals (Case Study 2.3). Some
have therefore had to sell or set free animals since the
earthquakes. Farming was indirectly affected by these
impacts on livestock not only because draft animals
21 The Asia Foundation and Democracy Resource Center Nepal
(2016). Aid and Recovery in Post-Earthquake Nepal: Independent
Impacts and Recovery Monitoring Phase 2 - Qualitative Field
Monitoring (February and March 2016). Kathmandu and
Bangkok: The Asia Foundation, pp. 63-64.
21


Recovery
were killed but also because there was less manure
available. One farmer in Barpak VDC, Gorkha, stated,
“Since my cattle died in the earthquake, I cannot
cultivate my land as I have no fertilizer and I have
given up cultivating.” “We used to rear cattle and goats.
But they were trapped and died in the earthquake and
now we have no dung to use in agriculture, so we gave
up farming,” added an old woman.
Case Study 2.3: A displaced Dalit is struggling to resume farming
Prem Bahadur Sarki’s house was fully damaged
during the earthquakes and his agricultural
land was damaged by cracks and is at a high risk
of landslides. He was displaced along with an-
other 40 households (20 of them Dalit families)
from his settlement in Prapcha VDC, Okhald-
hunga. The Dalit households were resettled by
the government in temporary shelters far away
from their settlement. This introduced some
tensions with the local community nearby and
also made farming difficult due to the lack of
shelters for seeds, harvests, and livestock, and
the long distance to their land.22
Prem Bahadur said he faced problems manag-
ing his livestock and fields from the temporary
shelter: “I was living in one place and my live-
stock were in another place. [...] I want to go
back to my own place but I cannot because the
area is prone to landslides.” To be closer to his
land and have more space for his cattle, Prem
Bahadur left the temporary shelter provided by
the government and moved to an upper caste
settlement closer to his land. There, he rented a
small plot of land to construct a new temporary
shelter for his family and a shed for his buffalo.
But this land was small and Prem Bahadur had
to rent yet another piece of land to collect grass
for his buffalo and firewood for cooking, further
adding to the financial burden imposed by the
destruction of his house and land.
Prem Bahadur continued to farm his own land
despite risks and being scared. “If I don’t cul-
tivate my land, I don’t have enough to eat. If I
cultivate, I risk my life because of landslides,”
he said. Indeed, during the heavy monsoon
rains, a landslide swept away whatever land
he had left. He pointed to the hill on the other
side and showed a small patch of land. “I had
planted maize with difficulty on my land but the
landslide swept away everything,” he explained.
Prem Bahadur is concerned that even with the
cash assistance provided by the government,
he may not be able to rebuild as he no longer
has any land. He said, “Where will I build my
house even if I receive money and will I be able
to receive money if I don’t build a foundation?”
Referring to his debts of over NPR 50,000 he
exclaimed, “I am in a state of despair, will the
government understand the plight of people
like us?”
Those facing particular difficulties
A number of factors determined whether
people were able to recover their livelihoods
or not, some of them unrelated to the earth-
quakes. Many pre-existing hardships were
exacerbated pushing those already poor
further into poverty, especially poor farmers
and Dalits.
While the initial impact of the earthquakes on live-
lihoods was major and widespread, only a limited
number of households faced a complete loss of their
livelihood. Several factors helped livelihoods recover
in the first year after the earthquakes. First, markets
reopened within the first few months after the earth-
quakes and businesses—with the exception of the
tourism sector—were able to resume operating, at least
to some extent. Second, farming, the most common
livelihood in the earthquake affected districts, gener-
ally resumed after the 2015 monsoon. And, third, the
diversification of incomes commonly practiced by the
22 See Case Study ‘Resettlement and caste - different responses of
displaced communities in Okhaldhunga’, The Asia Foundation
and Democracy Resource Center Nepal (2015). Aid and Recovery
in Post-Earthquake Nepal: Independent Impacts and Recovery
Monitoring Phase 1 - Qualitative Field Monitoring (June 2015).
Kathmandu and Bangkok: The Asia Foundation, p. 69.
22


Aid and Recovery in Post-Earthquake Nepal
majority of affected households meant that only very
few households lost all of their income sources.23 24 23
However, challenges remained and many of those
households who have been able to work again since
the earthquakes have simply found themselves contin-
uing to live in poverty, especially poor farmers. While
families have no choice other than to be resilient and
to cope with the impacts of the earthquake by working
hard, poverty is making full recovery hard for many.24
The 2015 earthquakes exacerbated hardships people
were already suffering from. As an old woman in
Okhaldhungsa said, “I was already poor, the earth-
quake pushed me further into poverty.”
This particularly affected Dalits. Generally being poor
and owning smaller or no land, Dalits have long had
to rely on traditional crafts or wage labor just to buy
enough food. With less diverse income opportunities
and assets than other groups, Dalits were still signifi-
Change of livelihoods
Very few have changed livelihoods since the
earthquakes.
Around 2% of people in all affected regions report that
they have changed their livelihoods since IRM-2.27
While the majority of these people have changed to
farming (70%), 14% have turned to their own business,
8% to daily wage work, 4% to relying on remittances,
and the remaining 4% to other income sources. The
majority of those who changed to farming in IRM-3
mention livestock farming as their main income source
in the earlier survey.28 Findings from the qualitative
research, on the other hand, suggest that wage labor
was becoming more common due to increasing
opportunities in the construction sector, with those
changing occupation turning away from farming or
crafts to wage labor.
cantly slower to recover despite increases in work and
wages for laborers.
Incomes were recovering both for people who were
in their own house and those in shelters. However,
those who were in their own house were more likely
than others to report that at least one income source
was not recovering.25 Those who moved from shelter
to home were less likely to report improvement if their
income sources are farming, daily wages, remittances,
or private salaries compared to others.26 But income
improvement was more likely if their income sources
are their own business, government salaries, rent
and livestock farming. These findings suggest that for
some, trade-offs are being made between investing in
housing or in their livelihoods. Those who had already
finished rebuilding their house were more likely than
others to have had an income source recover in the
three months prior to September 2016.
Many highlighted the lack of alternative opportunities
as the reason for the low rate of changes in livelihoods.
Farmers often pointed out that it is not easy to change
profession in the village, where there are no options to
earn money other than through agriculture, especially
in remote areas. As Sitamaya Tamang from Okhald-
hunga said: “Even if the earthquake damaged my
house, I don’t have a choice but to farm my land. I was
farming before the earthquake and I am farming now.”
Another resident from the same district speculated:
“People might change occupation if they have other
options but the VDC does not offer any alternative
economic activity, so people are compelled to go back
to agriculture and livestock farming because of the
lack of choices.”
23 In the districts visited, households have long relied on multiple
income sources in addition to small-scale agricultural production
or subsistence farming. The Asia Foundation and Democracy
Resource Center Nepal (2015). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Phase 1 - Qualitative Field Monitoring (June 2015). Kathmandu
and Bangkok: The Asia Foundation, pp. 85-93.
24 Most people in four of the districts studied are living in poverty.
In Ramechhap, Okhaldhunga, Gorkha, and Sindhupalchowk,
daily per capita incomes are between USD 2.60 and USD 3.00.
In Syangja, the average per capita income is USD 3.30 and in
Solukhumbu it is USD 5.10. See, UNDP Human Development In-
dex 2014: Nepal Human Development Report, 2014. UNDP/NPC.
Poverty is disparate within districts, but the poverty incidence is
between 5i.4%-82.2% in parts of Gorkha, Syangja, Okhaldhunga,
Ramechhap, and Solukhumbu, and half of Sindhupalchowk has
a poverty incidence between 42.9%-5i.4%. http://www.un.org.
np/node/10125
25 The findings hold if we look at results for where at least one in-
come source has not recovered (as people have multiple income
sources, some may have recovered while other have not). Sixteen
percent of people in their own house report that at least one
source has not recovered compared with 20% for those in shelters
on their own land and 31% for those in shelters on others’ land.
26 According to the IRM-2 and IRM-3 household panel dataset,
nearly 12% of people that were living in shelter in IRM-2 moved
to their own houses in IRM-3. This dataset was used to determine
if these individuals also report improvements in their income
sources.
27 IRM-3 weighted dataset.
28 IRM-2, IRM-2 household panel dataset, unweighted.
23


Recovery
2.4 Food
The need for food in all affected districts has
declined in IRM-3 compared to earlier re-
search rounds.
Compared to IRM-1 (June 2015), there was a 10 per-
centage point decline in people reporting food as one
of their most important immediate needs in IRM-2
(February-March 2016), and another 7 percentage
point decline from IRM-2 to IRM-3 (September
2016). Similarly, when asked about their most impor-
tant needs for the next three months, there was a 10
percentage point drop in the proportion of people re-
porting food between IRM-1 and IRM-2 and a further
4 point drop between IRM-2 and IRM-3. However,
nearly 10% of people continued to reported food as a
priority need both for immediate purposes and for the
next three months (Figure 2.8).
Figure 2.8: Food as a top immediate need and three month need (IRM-1, IRM-2, IRM-3, weighted)
Immediate food need H Need next 3 months
Food was reported as being more urgently
needed in severely impacted districts.
Individuals in severely impacted districts reported a
very high need for food in their households compared
to other districts. Only 3% or less in other impact cat-
egories mention food as one of their most important
immediate needs, compared to 26% of people in the
severely hit districts. Similarly, 28% in severely hit dis-
tricts mentioned food as a priority need for the three
months after September 2016, compared to only 2%
or less people in other district categories (Figure 2.9).
Amongst districts that were not severely hit, the pro-
portion of people prioritizing food as a current need
was highest in Okhaldhunga (8%), Solukhumbu, and
Bhaktapur (both 7%). In IRM-2, Solukhumbu had the
highest share of people reporting food as the most im-
portant immediate need.29 However, the stated need
for food has declined there and the districts with the
highest reported levels of food needs were all severely
hit ones. Gorkha now has the highest proportion of
people reporting food as a priority current need (32%)
but proportions were also high in every other severely
hit district with the partial exception of Ramechhap.
More remote and rural areas reported the
greatest need for food.
In more remote areas, 18% mentioned food as a priori-
ty immediate need, and 21% mentioned that it was the
most important need for the three months following
September 2016. Food was an immediate need for
13% and a need for the next three months for 14% of
people in remote areas. In contrast, only 4% or less
people in less remote wards mentioned food as one
of their most important needs in the immediate term
or for the next three months. Food need was nearly
seven times higher in rural areas than in urban areas.
29 Thirty-one percent of people in Solukhumbu said that food was
amongst their top two immediate needs in IRM-2. The Asia
Foundation (2016). Aid and Recovery in Post-Earthquake Nepal:
Independent Impacts and Recovery Monitoring Nepal Phase 2:
February-March 2016. Quantitative Report. Kathmandu and
Bangkok: The Asia Foundation, p. 79.
24


Aid and Recovery in Post-Earthquake Nepal
Figure 2.9: Food as a top immediate need and three month need - by district impact (IRM-3, weighted)
Food need immediate Food need next 3 months
Food needs differed among different popu-
lation groups with those with disabilities,
Janajatis, and from low castes more likely to
report needing food.
Food needs were much higher for those with disabilities
than those without. When asked about their priority
immediate needs, people with disabilities were twice
as likely to mention food as people without a disability.
Similarly, when asked about priority needs for the
next three months, 19% of people with a disability
mentioned food compared to only 10% without any
disability. High caste individuals were less likely to
mention a need for food compared to Janajatis or low
caste individuals. The stated need for food was slightly
higher among Janajatis than low caste individuals.
Food consumption has remained similar
between IRM-2 and IRM-3.
Most people said their food consumption stayed the
same since the end of the winter in February 2016.
Twenty-one percent said that food consumption
increased while 4% said it had decreased. These find-
ings were similar to those from IRM-2. There was a
significant drop in IRM-3 in the number of people
who reported increased consumption but also a small
decline in the number who said food consumption
had declined.
While relatively low numbers of people report decreas-
es in food consumption, some districts saw higher
numbers of people consuming less. In Sindhupalcho-
wk, 18% of people reported a decrease in consumption.
Other districts with a notable decrease in food con-
sumption were Ramechhap (8%), Okhaldhunga (8%),
and Lamjung (8%). However, people in every district
were more likely to report increased consumption
than decreases, with between one-quarter and around
one-third reporting increases in severely hit districts.
The data looking at changes in food consumption over
the last year show similar figures to the six-month
changes suggesting that recent improvements in
food consumption were not due to seasonal varia-
tion. Districts with the highest proportion of people
reporting decreases in year-on-year consumption were
Solukhumbu and Ramechhap (both 8%) and Okhald-
hunga and Lamjung (both 7%).
In the qualitative research, few mentioned food as
a priority need. However, some farmers reported
reduced yields due to earthquake impacts and other
factors such as water shortages, and malnutrition was
observed to be on the rise in some locations, especially
among children.
25


Recovery
2.6 Trauma and vulnerabilities
Psychological effects of the earthquakes
Many people were still suffering psychologi-
cally from the earthquakes.
Nineteen percent of people said someone in their
household still suffered; another 4% said someone in
the family had been suffering psychologically, but was
getting better. Trauma was widespread in the severely
hit districts, especially Sindhupalchowk (36%), along
with Okhaldhunga (34%). Yet, no direct correlation
between prevalence and the severity of earthquake
impacts was found. Notably, psychological impacts
were most widespread in the hit district of Syangja
(37%), the least affected district in the sample. This
may be because Syangja has received less attention
from aid providers, and presumably specialists in psy-
chosocial care, than other districts. However, Syangja
is a district that generally has high rates of suicide and
prevalence of depression.30 The likelihood of experi-
encing enduring psychological effects increases with
remoteness and was more prevalent in rural areas
than in urban ones.
Extreme fear and startling while sleeping
were the most common psychological effects.
Among those who reported a family member suffering
psychological effects from the earthquakes, 47% said
the family member has extreme fear and 38% said they
get startled while sleeping. Eleven percent mentioned
trouble sleeping and 4% nervousness.
The qualitative research suggests that psychological
distress from seeing one’s homes and belongings de-
stroyed and losing family members has made some
turn to alcohol. As Beg Bahadur Gurung from Barpak
VDC in Gorkha said, “Alcohol consumption is too high
after earthquake, some people have lost their lives al-
ready because of the excessive consumption of alcohol.”
Vulnerability
Landslides continued to be a common worry
and increased vulnerability.
Destruction of houses and damage to land, including
fissures and landslide risks, have been highlighted in
the previous rounds of IRM as the main issues expos-
ing people to vulnerable environments.31 Although
more were moving home by September 2016, land-
slides continued to be a risk in many areas, especially
during the monsoon. Syangja (46%), Sindhupalchowk
(35%), and Solukhumbu (31%) had the highest share
of respondents saying there was a landslide in their
area. The likelihood of a landslide increased sharply
with remoteness and landslides tended to occur more
in rural (18%) than in urban (5%) areas.
People in the severely hit (41%) and hit (50%) districts
were the most likely to worry about possible landslides
in their community with the onset of the monsoon.
Majorities in Sindhupalchowk (69%), Okhaldhunga
(52%), and Syangja (50%) were worried. Concerns
over possible landslides were also much more com-
mon in remote and more remote areas compared to
less remote areas. People in rural areas (30%) tend to
be far more worried about the possibility of monsoon
landslides than those in urban ones (4%). Concern
about landslides tracks well with actual landslide oc-
currences, with 85% of those who reported landslides
in their area having been worried about possible land-
slides once the monsoon started.
The displaced and those living in temporary
shelters remained among the most vulnerable
groups.
Many continue to live in shelters and are exposed to
some levels of risk and vulnerability ranging from ex-
posure to harsh weather conditions and illnesses to, in
the case of the displaced, tensions with the local com-
munities in their new settlements (see Chapter 5), and
uncertainties about long-term settlement solutions.
The fact that some of those in shelters were returning
to damaged houses or landslide-prone land without
repairs or land assessments having been conducted
only increased their vulnerability.
30 This was frequently mentioned by informants for the qualitative
research and Syangja has the second highest suicide rates in
Nepal, after Ilam.
31 The qualitative research observed that most of the people
displaced from their land and house were living in temporary
shelters on public or rented land. People were observed to
generally stay close to their original homes. However, where
entire settlements were affected by damage to land or high risks
of landslides, local communities had to settle elsewhere in the
VDC or even beyond.
26


Aid and Recovery in Post-Earthquake Nepal
Photo: Nayan Pokharel
Geological landslide assessments remained
important to assess risks and determine long-
term resettlement for the displaced.
Given the prevalence of landslides and landslide
risks—both earthquake and monsoon related—geo-
logical land surveys remained a major need for some
areas. This has been repeatedly highlighted by the
IRM research.32 Some communities have now returned
to land with landslide risks while others whose land
was heavily damaged remain displaced and uncertain
whether they can return or where they will be resettled
in the long term if their land is unsafe. Some displaced
people received the cash grant for reconstruction, but
they were not sure if they could construct their house
on damaged or landslide-prone land. As a Dalit woman
in Barpak, Gorkha, said, “We requested the top leaders
of all parties that a geological survey be conducted so
that we can decide whether to build or not to build a
house there. If the survey says it is unsafe, the govern-
ment has to give us new land.”
People in remote areas continued to be more
vulnerable facing greater obstacles to access-
ing aid and rebuilding their houses.
As the survey data showed, people in remote areas have
generally been at a disadvantage compared to those in
less remote areas. For example, they were more likely
to have landslides, to be in need of food, or to have
to pay more for the transportation of construction
materials. The qualitative research highlights that
those in remote places also found it difficult to access
cash grants, due to longer travel time and higher costs
to reach locations where the required documents are
issued and the cash grants are disbursed via banks.
Yet, not all remote areas were equally disadvantaged.
Areas in Solukhumbu, for example, have received
more attention and assistance than remote parts of
Okhaldhunga. Not surprisingly, areas without road
access were the most disadvantaged.
Inequality and prevalent forms of exclusion
and discrimination negatively affect the
recovery of marginalized groups especially
of Dalits ivho stood out as a highly vulnerable
group in IRM-3.
Earthquake impacts observed across affected dis-
tricts were not experienced equally by all segments of
society. As time passes, it was becoming clearer how
structural inequalities and prevalent forms of exclu-
sion and discrimination negatively affect the recovery
of marginalized groups. This was predicted by the
IRM-1 report and the second round of research began
to observe the implications of this. At the time of IRM-
3, Dalits, marginalized ethnic groups, the poor and
economically disadvantaged, and the landless33 were
32 The government has since begun to conduct geological risk
assessments in many areas.
33 The landless and those living on public or guthi land (trust land,
a form of community-owned land) faced particular obstacles and
delays in receiving cash assistance as they could not sign cash
grant agreements at the time of research due to their lack of land
ownership certificates. Special provisions have since been made
to make it possible for these groups to receive reconstruction
cash assistance. See, The Procedure for the Reconstruction Grant
Distribution for Private Houses Damaged by Earthquake 2073
(2016) http://nra.gov.np/download/details/187
27


Recovery
more vulnerable. Lower and more unstable incomes,
fewer assets, limited access to formal sources of credit,
and owning little land or living with precarious land
arrangements were reported as drivers of the higher
level of vulnerability of marginalized groups. This
particularly affected Dalits.
Dalits faced greater obstacles to recovering their
livelihoods, generally relying on unstable sources of
income, and on finding additional financial and other
resources for rebuilding, including accessing loans. In
addition, most Dalits have smaller landholdings and
are often not allowed to use community and public
lands, which further limits their access to resources.
The historical structural marginalization of Dalits from
state and financial institutions is compounding the
problem. Dalits face barriers to effective participation
in decision-making processes. “No one listens to Dalits’
real concerns. Even the few Dalits in decision-making
bodies are used as tokens by the political parties to
serve their own interests. Even at the local level, our
voices are not acknowledged and we have no decision-
making power,” said a Dalit in Solukhumbu. While 1/
NGOs and others have provided special assistance to
Dalits in many areas, no comprehensive efforts exist
to counter the marginalization and vulnerability of
Dalits through earthquake-related recovery schemes.
Researchers observed one case of Dalits who were
unable to cope with their situation. In Dudhkunda
municipality in Solukhumbu, a Dalit couple committed
suicide due to severe financial stress, leaving their four
children orphans and destitute. Although their debts
were only partially related to the earthquake, it was
reported that the lack of assets, including land, and
income sources increased the couple’s debt burden.
Women, children, and the elderly continued
to be seen as particularly vulnerable groups
in most wards.
Women, children, and the elderly were considered
to be vulnerable. Qualitative findings show that they
were seen as being more vulnerable to health and
safety threats, especially in shelters. Children were
also reportedly more at risk of malnutrition. Women
faced risks of gender-based violence and trafficking.
Violence against women and girls was reported to have
increased after the earthquake in some districts al-
though no precise data is available and it is difficult to
link this to the earthquakes as gender-based violence
is generally common yet often under-reported. The
voices of women, children, and the elderly were also
rarely heard and included in decision-making, mean-
ing that their particular needs are rarely dealt with.
28


Aid and Recovery in Post-Earthquake Nepal
Chapter 3
Photo: Binu Sharma
3.1 Aid delivery
Types and coverage of aid
The coverage of aid has declined massively
since IRM-2 ivas conducted in March 2016.
By September, when IRM-3 was conducted, only 15%
of respondents said they had received any type of aid,
including material and cash support, since the end
of the winter season.34 This is a 39 percentage point
drop in the share of respondents reporting receiving
any aid compared to the six months prior to IRM-2
when 54% had received aid. Nearly everyone (96%)
said they received aid in IRM-1 in the weeks after the
earthquakes.
The decline in aid was large in districts of every level
of earthquake impact (Figure 3.1). Between IRM-1
and IRM-2, aid coverage dropped substantially in
the crisis hit districts (which include Kathmandu and
Bhaktapur) and the hit district of Syangja. There was
also a large drop in the hit with heavy losses districts
but two-in-three people there were still receiving aid
at the time of IRM-2. There was a very slight drop in
aid coverage in the severely hit districts. In contrast,
between IRM-2 and IRM-3, aid coverage continued to
plunge in the crisis hit, hit with heavy losses, and hit
districts, but also dropped steeply in the most-affected
severely hit districts. While people in the severely hit
districts were the most likely to have received aid
since the end of the winter (26% received aid) this is
a decline from 98% in IRM-2.
The drop in aid coverage was most pronounced in the
severely hit districts of Dhading (a 90 point drop),
Nuwakot (84 points), and Ramechhap (76 points),
along with the less affected Solukhumbu (79 point
drop) - Table 3.1. Aid coverage was wider in Gorkha
than elsewhere with a majority of people saying they
received aid since the end of the winter.35 Aid coverage
in Solukhumbu was particularly expansive in IRM-2
compared to other similarly impacted districts but
there has been a significant drop in aid since then.36
34 Survey respondents are first asked whether they received aid.
They are then given a list of different types of aid and asked
whether they received any of them. From this, we can determine
whether people received aid of any type or not. In September
2016, aid from the government focused largely on the housing
reconstruction program but few had received cash under this
program when fieldwork was conducted in September 2016. Many
more have received such cash grants since then (see Annex A).
35 Gorkha now has a higher share of people living in their own house
than other severely hit districts. See Chapter 2.
36 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 2: February-March 2016. Synthesis Report. Kath-
mandu and Bangkok: The Asia Foundation, p.25.
29


Aid and Housing Reconstruction Cash Grants
There appears to have been no significant distribution of any type of aid in Lamjung or Bhaktapur since the
end of the winter season.
Figure 3.1: Proportion of people receiving aid - by district impact (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-l â–  IRM-2 â–  IRM-3
Table 3.1: Proportion of people receiving aid - by district impact
and district (IRM-1, IRM-2, IRM-3, weighted)
IRM-1 IRM-2 IRM-3 Decline in coverage between IRM-1 and IRM-2 (percentage points) Decline in coverage between IRM-2 and IRM-3 (percentage points)
Severely hit 100% 98% 26% 2% 72%
Dhading 100% 97% 7% 3% 90%
Gorkha 100% 97% 56% 3% 41%
Nuwakot 100% 99% 15% 1% 84%
Ramechhap 100% 97% 21% 3% 76%
Sindhupalchowk 100% 100% 32% 0% 68%
Crisis hit 92% 30% 11% 62% 19%
Bhaktapur 100% 55% 0% 45% 55%
Kathmandu 91% 23% 11% 68% 12%
Okhaldhunga 100% 76% 34% 24% 42%
Hit with heavy losses 100% 65% 6% 35% 59%
Lamjung 100% 47% 0% 53% 47%
Solukhumbu 100% 95% 16% 5% 79%
Hit 100% 30% 5% 70% 25%
Syangja 100% 30% 5% 70% 25%
All districts 96% 54% 15% 42% 39%
30


Aid and Recovery in Post-Earthquake Nepal
Table 3.2: Proportion of people receiving different
types of aid (IRM-1, IRM-2, IRM-3, weighted)
IRM-1 IRM-2 IRM-3
Shelter
Tent 1% 2% 1%
Tarps 45% 31% 2%
Corrugated iron sheets 6% 16% 1%
Reconstruction materials) - 4% 0%
Cash
Non-government 20% 10% 2%
Government 48% 8%
Livelihoods
Farm implements — 4% 1%
Livestock - 0% 0%
Other
Food 37% 28% 2%
Medical aid 3% 4% 0%
Sanitation package/kit 8% 11% 1%
Blankets 11% 24% 3%
Warm clothes 1% 2% 2%
Solar 0% 3% 0%
Kitchen set 4% 1% 1%
Mattress 1% - -
* In IRM-1, cash was not separated into government
and non-government cash and clothes were not
specified as being warm clothes. Reconstruction
material, farm implements and livestock were not
included nor mentioned by respondents in IRM-1.
Mattresses were not included nor mentioned by
respondents in IRM-2 or IRM-3.
The massive drop in the coverage of aid ivas
true for different types of assistance including
relief material aid, and cash support.
In terms of shelter, the distribution of tarps and CGI
has fallen steeply since IRM-2, unsurprising given
that the focus was firmly on reconstruction rather
than emergency support. However, this did not lead
to an increase in the provision of materials for recon-
struction. In fact, while some people received recon-
struction materials in IRM-2, no-one did in IRM-3.
The approach of the government and major donors to
reconstruction has largely focused on providing cash
for reconstruction. However, the number of people
receiving cash in the six months before the IRM-3
survey was conducted dropped significantly since the
period preceding IRM-2. Forty-eight percent of people
in IRM-2 had received cash from the government37 but
only 8% did in IRM-3. The distribution of food aid has
also fallen massively: from 37% in IRM-1 receiving
food to 28% in IRM-2 and just 2% in IRM-3.
The coverage of aid also decreased in severely
hit districts.
From the earthquakes, the severely hit districts re-
ceived more of most types of aid than other areas. In
IRM-3, too, the severely hit districts got more aid.
Severely hit districts also generally had a higher pres-
ence of non-governmental actors involved in recovery
efforts.38 However, aid coverage in these districts has
shrunk dramatically for every type of aid. For instance,
the share receiving cash from the government in the
severely hit districts is 15% compared to 91% in IRM-
2, and the share receiving cash from non-government
organizations dropped from 21% to 3%. 2% got farm
implements compared to 13% in IRM-2, and 3% food
compared to 68% in IRM-2. The provision of relief
materials in the form of tarps, CGI, blankets, clothes,
and sanitation kits also dropped significantly in se-
verely hit districts.
Solukhumbu was receiving comparatively more aid
than other districts, as already reported in IRM-2,
mostly from individual private donors. Thirty-two
organizations were registered as conducting earth-
quake recovery schemes at the District Development
Committee (DDC) in Solukhumbu. Yet, researchers
only encountered one of these organizations in the
VDCs visited, which was rebuilding damaged school
buildings.
The provision ofcash has play ed a role in de-
termining whether people were able to return
to their house.
Cash from government and non-government providers
appears to have played an important role in allowing
people to repair or rebuild houses. Individuals who
have received cash from non-government agencies
were 8 percentage points more likely to transition from
37 During the monsoon of 2015, the government provided NPR
30,000 for funeral costs for those households who lost a member
during the earthquake, NPR 15,000 for households with ‘red
cards’ (those whose house was ‘fully damaged’) to build temporary
shelters, and NPR 3,000 for households with ‘yellow cards’ (those
with ‘partially damaged’ houses). This was followed by the winter
relief grants of NPR 10,000 distributed between October 2015 and
March 2016. See details in ibid., pp. 3-5.
38 The qualitative research revealed that the number of non-
government organizations and activities remained higher in
Gorkha and Sindhupalchowk compared to other districts,
as in previous rounds of research. For instance, 24 INGOs,
19 NGOs, and seven UN agencies were working on recovery
in Sindhupalchowk, while Syangja had no registered I/NGO
support. In the VDCs visited in Solukhumbu, Ramechhap, and
Okhaldhunga, the number of programs supported by I/NGOs
were none to four. Only one NGO was found in Doramba VDC in
Ramechhap and no NGOs were found to be present in the other
two VDCs visited in the district even though it is listed in the same
damage category (severely hit) as Sindhupakchowk and Gorkha.
39 This analysis is based on the panel dataset of 1,470 individuals
who were interviewed in all three rounds of the survey.
31


Aid and Housing Reconstruction Cash Grants
shelters to their own houses between IRM-1 and IRM-
3.39 The results for government cash grants were even
stronger. Twenty-six percent who were in temporary
shelters who received cash from the government
moved into their own house compared to 11% of those
who did not receive government cash (Figure 3.2).
Figure 3.2: Proportions of people receiving and not receiving cash who moved from shelter to home -
by government vs. non-government cash (IRM-1, IRM-2, IRM-3 household panel, unweighted)
Aid providers
The government remained the most prom-
inent aid provider, followed by INGOs and
NGOs. The share of people receiving aid from
individual donors declined significantly.
As Figure 3.3 shows, the Nepal government, including
VDC and municipalities, was the most prominent
provider of assistance: 60% of those receiving aid in
IRM-3 received it from the government.40 The share
mentioning these bodies, however, declined compared
to previous surveys (78% amongst those who received
aid in IRM-1 and 90% in IRM-2). The second most
common provider was INGOs (22% of those who
received aid receiving assistance from INGOs), similar
levels as in previous surveys. NGOs were the third
most common provider (16%, down from 36% in IRM-
2). Other major donors in previous waves of the survey
saw their prominence decline. Individual donors have
declined sharply from 15% during the early response
period (IRM-1) to 7% in IRM-2 and just 1% in IRM-3.
There was significant variation between districts in who
was providing aid. In most of the severely hit districts,
along with Solukhumbu, Syangja, and Kathmandu, the
vast majority of those receiving aid were receiving it
from the government. However, the government was
much less important in Sindhupalchowk—despite
distribution of reconstruction cash grants having
40 Because people may have received aid from multiple providers,
numbers do not add up to 100%.
started there—and in Okhaldhunga. In both districts,
INGOs were covering many more people than the
government.
Government and non-government providers
offered different types of assistance. The
governmentfocused on the distribution ofcash
grants while I/NGOs mostly provided ‘soft’
forms of aid through trainings, awareness
raising, and technical assistance.
Survey data show that the government was the major
source of temporary shelter items (Table 3.3). Among the
1% of people who received tents, the government (80%)
is the most common provider followed by individuals,
local government-affiliated people and organizations,
and NGOs (20% each). The government was also the
main provider of tarps (45%). Provision of CGI was
slightly more common by INGOs than by the government
(51% to 47%). Forty-four percent of people who received
cash from a non-governmental source said that INGOs
provided cash grants, slightly more than the 37% who
said that NGOs provided cash. The government was
listed as the source for some non-government cash as
well, which could be due to cash from non-governmental
sources ultimately being disbursed from a government
body. The government was also the predominant
provider of food, sanitation packages, blankets, and
warm clothes. Most of the farming implements (89%)
and kitchen sets (90%) were provided by INGOs.
However, it should be noted again that very few people
were receiving any of these types of aid.
32


Aid and Recovery in Post-Earthquake Nepal
Figure 3.3: Sources of aid amongst those who received aid (IRM-1, IRM-2, IRM-3, weighted)
I IRM-1 â–  IRM-2 â–  IRM-3
Table 3.3: Type of aid provided - by source (IRM-3, weighted)41
Nepal Government/ VDC/ municipality LGCDP/WCF/ CAC/SM* Political parties Red Cross Individuals Business groups NGOs INGOs Donors (except UN) Other countries Don’t know
Tent 80% 20% 0% 0% 20% 0% 20% 0% 0% 0% 40%
Tarps 45% 6% 5% 12% 5% 0% 16% 3% 0% 21% 17%
Corrugated iron sheets 47% 11% 3% 3% 13% 0% 19% 51% 1% 1% 7%
Food aid 58% 6% 0% 14% 5% 19% 18% 14% 0% 5% 8%
Cash: non-government 31% 0% 0% 12% 0% 0% 37% 44% 3% 5% 8%
Cash: government 100% 2% 0% 3% 2% 0% 13% 11% 0% 2% 5%
Sanitation package 32% 1% 0% 2% 0% 0% 6% 27% 6% 0% 34%
Farm implements 8% 0% 0% 0% 0% 0% 9% 89% 0% 0% 2%
Blankets 77% 6% 0% 2% 5% 0% 28% 25% 0% 0% 10%
Warm clothes 63% 1% 0% 3% 2% 0% 24% 52% 1% 0% 7%
Kitchen set 48% 2% 0% 4% 0% 0% 6% 90% 0% 0% 5%
*Local Governance and Community Development Prog ram me/Ward Citizen Forum/
Community Awareness Center, Social Mobilizes
Findings from the qualitative research help us further
analyze this data. At the time of research, the govern-
ment was focusing on the provision of reconstruction
cash grants to rebuild houses. Some limited livelihoods
assistance was also provided by the government, es-
pecially in Solukhumbu and Syangja, the two of the
six districts visited for the qualitative research where
reconstruction cash grants were not yet available to
earthquake victims. Livelihoods support included the
distribution of seeds, livestock, and farming tools. Be-
41 Percentages add up to more than too as multiple responses were
allowed.
33


Aid and Housing Reconstruction Cash Grants
tween IRM-2 and IRM-3 the government also contin-
ued to distribute winter assistance in areas where there
had been delays.42 The government was not found to be
distributing materials other than livelihoods support
in the qualitative research.
The qualitative research also found that while some
non-governmental organizations were providing ‘hard’
material assistance, most of their projects at the local
level were focusing on ‘soft’ forms assistance such as
sanitation and hygiene awareness, livelihood support
programs, trainings, and technical assistance. Some
of these programs, especially livelihoods support
programs, have continued since before the earth-
quakes. New types of ‘soft’ assistance were trainings
related to reconstruction, psychosocial counseling,
and disaster awareness. Overall, I/NGO support for
reconstruction was found to have increased, albeit in
the form of ‘soft’ assistance. In the VDCs visited for
the qualitative research, there were only few exam-
ples of non-governmental actors and private donors
directly building, or planning to build, private houses
for earthquake victims. Some I/NGOs however were
helping to directly rebuild infrastructure such as
water and irrigation systems, schools, health centers,
or roads. It should be noted that the NRA requested
that I/NGOs focus on ‘software’, especially technical
assistance. Further, many I/NGOs were still waiting
for their ‘hardware’ projects to be approved by the
government at the time of research.
Experience of aid among different population groups
People in more remote areas were more likely
to have received aid but some extremely re-
mote areas could not be reached during the
monsoon.
Aid between the end of the winter season and Sep-
tember 2016 was more likely to reach more remote
areas than urban centers. These areas were more
likely to have received aid in IRM-i and IRM-2 as well.
Thirty-three percent of those in more remote wards
(between one and six hours away using the regular
means of getting to the district headquarters) had
received aid during IRM-3.
More remote and rural areas could have received
higher levels of assistance in part because they make
up a higher share (69% more remote, 47% rural) of the
severely hit districts, where more aid was given (26%
received aid in the severely hit districts). However,
even among those whose house was completely
destroyed by the earthquake, those in more remote
areas (37%) are more likely to have received aid than
people in remote (21%) and less remote (22%) places.
Twenty-five percent of those in rural areas whose
houses were completely destroyed got aid compared
to 18% in urban areas.
While, those in rural and more remote areas received
more types of assistance, people in more remote areas
were less likely to receive cash from the government
(40%) than those living in remote and less remote
places (59% in both). In contrast, the likelihood
of receiving cash from non-governmental sources
increases with remoteness. Of those who received
aid in IRM-3, people in more remote areas were less
likely than others to have received aid through the
government and NGOs. On the other hand, six in
10 respondents living in more remote areas report
receiving aid from INGOs, compared to just 2% in less
remote and 18% in remote areas.
The qualitative research suggests that very remote
wards received less assistance compared to more
accessible ones between IRM-2 and IRM3, and espe-
cially during the monsoon months. This was primarily
because of transportation problems due to monsoon
rains with air transport being too expensive. For exam-
ple, remote and inaccessible VDCs like Goli, Bhakanje,
and Chaulakharka, in the eastern part of Solukhumbu
district bordering Ramechhap and Dolakha, were par-
ticularly hard hit by the earthquakes. Yet, according
to data from the district offices, these VDCs received
less relief and aid due to their extreme remoteness.
In Gorkha and Sindhupalchowk, too, district officials
said that very remote areas were largely unattended
to by I/NGOs. Respondents generally agreed that this
disparity was mainly due to poor access and was not
deliberate.
Disability, caste, and gender did not appear
to determine access to aid but there were
systematic differences in access to certain
types of aid and access to aid from different
providers across groups.
42 In February-March 2016 (IRM-2), some VDCs had not yet re-
ceived the government’s winterization support of NPR 10,000.
At the time of IRM-2,64 households in Syaule VDC in Sindhupal-
chowk, for example, had not received the winter cash grant due
to insufficient budget. The budget was released only in August
2016 by the District Disaster Relief Committee (DDRC) for 41
households considered eligible. This money was then distributed
equally among the 64 households initially considered eligible in
the VDC, amounting to around NRP 6,400 for each household.
34


Aid and Recovery in Post-Earthquake Nepal
As was the case in IRM-1 and IRM-2, similar shares
of men (17%) and women (13%) received aid in IRM-
3. Among those who receive aid, men were slightly
more likely than women to have received cash from
non-governmental sources (17% to 15%) and from the
government (58% to 53%). Women were more likely
to have received tents, sanitation packages, and warm
clothes.
There were no major differences in the likelihood of
receiving aid for those with a disability (17% received
aid) and those without (15%). However, 65% of people
with a disability reported that their houses are fully
damaged, which is a criterion for receiving many types
of aid, compared to 52% of those without a disability.
Among those who received aid, those without a
disability were more likely than those with one to
receive most types of assistance. However, those with a
disability were slightly more likely than those without
one to get government cash (60% to 55%). Those with
a disability were less likely than those without to have
received aid from the government (63% to 54%) and
NGOs (8% to 15%).
At a time when levels of assistance provided was low,
similar shares across caste and ethnic groups reported
receiving aid. In IRM-1 and IRM-2, Janajatis and
those belonging to lower castes were more likely than
those belonging to higher caste groups to receive aid
(Figure 3.4). Of those who received assistance in IRM-
3, Janajatis were less likely than those belonging to
high or low caste groups to report receiving aid from
the government. Janajatis were more likely to be
served by INGOs while NGOs reached those belonging
to lower castes. Janajatis were more likely to receive
cash from non-governmental sources, but less likely
to get it from the government.
Figure 3.4: Proportion who received aid - by caste (IRM-1, IRM-2, IRM-3, weighted)
Low caste H Janajati H High caste
Housing damage determined volumes and
types of assistance received.
Unsurprisingly, those whose houses were completely
destroyed (23%) were more likely to have received
aid in IRM-3 than those whose houses were badly
damaged (13%), those whose houses need minor
repairs to make it habitable (9%), or whose houses
were not damaged (2%).43 Among those who received
aid, government cash in IRM-3 went to people whose
houses were completely destroyed. However, those
43 In IRM-1 and IRM-2, nearly everyone whose house was completely
destroyed received aid. Eighty percent of those whose houses were
badly damaged received aid in IRM-1 and 74% in IRM-2.
with badly damaged houses (45%) were more likely
than those with completely destroyed houses (13%) to
get cash from non-governmental sources. Tarps and
blankets went to people with lower levels of housing
damage.
Those with lower incomes were less likely to
receive government assistance.
Among those who received some form of assistance,
those with higher incomes tended to be more likely
to receive cash from the government while those with
lower incomes were more likely to have received cash
from non-governmental sources. Though majorities
across income groups mention the government as a
source of aid, those in the middle and high income
35


Aid and Housing Reconstruction Cash Grants
groups were more likely to do so than those in the
lower income group. People in the lower and high
income groups on the other hand were more likely
to have received aid from NGOs. The likelihood of
having received aid from INGOs decreased with rising
income (39% low income, 32% middle income, 21%
high income).
3.2 Needs
Cash and construction materials were the
most widely cited current need.
The top five current immediate needs identified by
survey respondents were cash (59% identified it as a
top three need), items to reconstruct houses (30%),
CGI (11%), rice, wheat, and maize (10%), and livestock
(9%). Fewer mentioned clean drinking water, clean
water for household use, medical aid, warm clothes,
sugar, salt and spices, farm implements, lentils,
blankets, tarps or sanitary materials (each 2% or less)
- Table 3.4.
Nearly nine in 10 in the severely hit districts said cash
was a current priority need. Those in Okhaldhunga
(92%), a crisis hit district, and Solukhumbu (80%), a
hit with heavy losses district, also mentioned cash more
often than people in other districts. Reconstruction
material was mentioned most frequently in Nuwakot
(81%) and Sindhupalchowk (70%). Respondents in
Kathmandu tended to mention livestock (19%) and
those in Solukhumbu considered farm implements
(15%) a priority current need.
Table 3.4: Top five current needs - by district impact and district (IRM-3, weighted)
Cash Items to reconstruct house Corrugated iron sheet Rice, Wheat, Maize Livestock
Severely hit 93% 67% 27% 26% 1%
Dhading 95% 61% 22% 29% 1%
Gorkha 88% 65% 21% 32% 1%
Nuwakot 95% 81% 56% 26% 3%
Ramechhap 97% 57% 17% 14% 2%
Sindhupalchowk 89% 70% 20% 23% 1%
Crisis 43% 9% 2% 2% 16%
Bhaktapur 60% 36% 0% 7% 3%
Kathmandu 36% 2% 0% 1% 19%
Okhaldhunga 92% 41% 29% 8% 0%
Hit with heavy losses 66% 37% 20% 3% 2%
Lamjung 58% 29% 13% 1% 0%
Solukhumbu 80% 51% 34% 7% 4%
Hit 25% 13% 5% 2% 1%
Syangja 25% 13% 5% 2% 1%
All districts 59% 30% 11% 10% 9%
Findings from the qualitative research confirm that
the reconstruction of houses and related support
was the priority for communities. Some other needs
mentioned were directly related to reconstruction such
as cash grants, soft loans, demolishing of old houses,
and availability of reconstruction materials such as
cement, sand, iron rod, wood, etc. Communities also
mentioned the need to improve road conditions to
transport reconstruction materials as necessary. Some
needs mentioned were specific to certain district or
VDCs. Citizens from Syangja and Solukhumbu, where
the CBS assessment had not yet been conducted, said
their priority needs were the CBS assessment and clear
information on the timeline and implementation of the
government’s reconstruction assistance.
The qualitative research also gives insight into the
wide variety of local needs. Improvement of basic
services, reconstruction of structures other than
houses, and psychosocial needs were still cited
36


Aid and Recovery in Post-Earthquake Nepal
as needs, but seen as comparatively less urgent
compared to IRM-1 and IRM-2. The same was true
for geological land assessments. Drinking water and
irrigation needs were heavily featured in IRM-2 as
the research was conducted in February/March (the
dry season) and less frequently mentioned in IRM-
3 (at the end of a good monsoon season), although
they remained important needs in many locations. In
contrast, citizens and officials interviewed in IRM-3
frequently mentioned the improvement of roads as
urgent because they had been destroyed by monsoon
rains and people had to travel to access cash grants via
banks and transport construction materials.
Needs have changed over time.
In all three surveys, respondents were asked to name
the most important current needs for them and their
household and what they anticipated would be needed
the most in three months. Comparing current and
future needs in each of the three survey waves allows
for an assessment of how needs have evolved over time,
shown in Figure 3.5. The share saying cash is the most
important need was at its highest at the time when
IRM-3 was conducted (59%). Reflecting immediate
food and shelter needs right after the earthquake,
the other two items mentioned most often as current
needs in IRM-1 were CGI sheets (37%) and rice, maize,
and lentils (27%). Both have declined in importance
for people although the amount prioritizing CGI has
risen sharply in IRM-3.
These top five needs were all expressed more com-
monly in the severely hit districts. Over time, more
people said cash was a need, with nearly everyone
(93%) prioritizing it in IRM-3. The share mentioning
rice, wheat, and maize declined sharply, but a quarter
of those in severely hit districts mentioned it as an im-
mediate need in IRM-3. Although fewer mentioned it
as a current need in IRM-3 (6%), clean drinking water
has also been consistently identified in the severely hit
districts. Shelter needs grew in IRM-3. Though the
projected need for construction materials declined in
IRM-2, it had grown by 39 percentage points at the
time IRM-3 was conducted, with 67% saying it is a
priority current need.
Figure 3.5: Changes in priority needs - IRM-1 and IRM-2 current and future needs,
IRM-3 current need (IRM-1, IRM-2, IRM-3, weighted)44
Cash H Corrugated iron sheet H Livestock
Items to reconstruct house H Rice, Wheat, Maize
The drop in aid coverage did not correspond
with any declining demandfor aid.
The share saying they need relief material at present
or in the next three months rose in the period between
IRM-2 and IRM-3. This was the case in every district
except the least affected district of Syangja. Rising
demand for aid suggests that people were realizing
that recovery has not been as speedy as they initially
thought it would be. In the severely hit districts, almost
everyone expressed needing aid now or in the future
in IRM-1 and IRM-2. In the crisis hit districts, 74%
in IRM-1 projected not needing aid in the future. By
IRM-2, this had declined to 60%. The share of people
holding this view slid further in IRM-3 (42% for both
44 Reconstruction materials and livestock were not included in
IRM-1.
37


Aid and Housing Reconstruction Cash Grants
current and future needs). Half of those residing in the
hit with heavy losses districts said they did not need
relief material in the future in IRM-1. By IRM-2, only
34% said they did not need aid at present. Though they
were more likely to say aid would not be needed in the
future (48%) in IRM-2, by IRM-3 only 29% said so of
the current situation and in the future.
The proportion of people who said they need no aid
now or in the next months was particularly low in all
of the severely hit districts along with Okhaldhunga
and Solukhumbu. Elsewhere, the proportion of people
saying they do not need aid any more was much
higher, ranging from 35% in Bhaktapur to 74% in
Syangja. However, in every district the share of people
saying they do not need aid was much lower than the
proportion of people who did not receive aid.
As time passes, the gap between needs and aid
provided seems to be increasing.
Aid provided has not fitted well with needs, in large
part because the coverage of aid was so low in IRM-3.
Looking at current needs mentioned in IRM-3, and
whether these items have been received since the
winter, shows the mismatch. Among those mentioning
cash as a current need, only 11% received it from the
government and 4% from non-governmental sources.
Among those who mention a staple food item as a
priority need, only 4% received any type of food aid.
Only 2% of those who say they need it received CGI
sheets. One percent of those who say they need them
received warm clothes. Of all those who mentioned
items to reconstruct houses, livestock, medical aid,
sanitary products and tents, none report having
received such items.
There seemed to be no shared understanding
and little coordination at the local level to
identify and prioritize needs.
Local government offices did not systematically
identify and record needs in communities nor coor-
dinate to facilitate a shared understanding of needs.
After the early weeks after the earthquakes, VDCs or
districts did not officially record and identify local
needs. Non-government organizations were some-
times conducting needs assessments before launching
their programs but local stakeholders said that most
non-governmental organizations did not conduct such
assessments. Where INGOs conducted needs assess-
ments, these were often limited to just the sector the
program works in or limited to a number of VDCs in
the district. This meant that there was no systematic
identification of needs or plan to address needs, nor
any shared understanding of needs between differ-
ent government offices or between government and
non-governmental organizations at the local level.
3.3 Housing reconstruction cash grants
Damage assessments
Generally people’s housing damage matches
their classification in the CBS damage as-
sessment. However, some misclassification
seemed to have taken place.
A series of damage assessments were conducted by
the government to decide on who should receive
beneficiary cards that would give them access to
various government cash grants.45 Respondents’ self-
classification of housing damage closely mirrored how
people’s houses were reportedly assessed in the most
recent damage assessment (the CBS assessment) but
the results suggest that some misclassification may
have taken place. Among respondents whose house
was classified as fully damaged, 91% said that their
house was completely destroyed while 1% said it was
not damaged at all (Table 3.5). Eighty-five percent
of people whose house was classified as partially
damaged said their house was impacted but not
destroyed by the earthquake. However, 8% of this
group said their house was completely destroyed and
another 7% said it was not damaged. Three percent of
those whose house was classified as not being damaged
said their house was completely destroyed and another
3% said it was badly damaged.
45 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 2 - Quantitative Survey (February and march
2016). Kathmandu and Bangkok: The Asia Foundation, and The
Asia Foundation and Democracy Resource Center Nepal (2016).
Nepal Government Distribution of Reconstruction Cash Grants
for Private Houses: IRM - Thematic Study (November 2016).
Kathmandu and Bangkok: The Asia Foundation.
38


Aid and Recovery in Post-Earthquake Nepal
Table 3.5: Housing classification in most recent damage assessment -
by self-reported housing damage (IRM-3, weighted)
Housing classification in the most recent damage assessment |
Fully damaged Partially damaged Normal/not damaged Don’t know
Completely destroyed 91% 8% 3% 23%
Self-reported levels of housing damage Badly damaged (needs major repair to live in) 6% 42% 3% 15%
Habitable (but needs minor repair) 2% 43% 36% 36%
Not damaged 1% 7% 58% 26%
Satisfaction with the official damage classifi-
cation was generally low.
In the severely hit districts, more people were satisfied
with the classification of their house than in IRM-2
but slightly more are also unsatisfied (Table 3.6).46
In contrast, in every other district, except Bhaktapur,
fewer people were satisfied than before and more
people are dissatisfied.
Table 3.6: Satisfaction with official damage classification -
by district and district impact (IRM-2, IRM-3 household panel, unweighted)
| IRM-2 | | IRM-3 |
Satisfied Unsatisfied Satisfied Unsatisfied ;
Severely hit 85% 4% 94% 6%
Dhading 57% 3% 91% 6%
Gorkha 94% 5% 94% 6%
Nuwakot 95% 3% 95% 5%
Ramechhap 79% 9% 92% 7%
Sindhupalchowk 95% 1% 94% 5%
Crisis hit 80% 14% 78% 17%
Bhaktapur 80% 11% 88% 11%
Kathmandu 84% 7% 66% 19%
Okhaldhunga 79% 19% 77% 20%
Hit with heavy losses 70% 24% 54% 42%
Lamjung 52% 38% 45% 53%
Solukhumbu 92% 7% 65% 28%
Hit 84% 14% 68% 31%
Syangja 84% 14% 68% 31%
All districts 82% 10% 82% 15%
Red = decrease in satisfaction/dissatisfaction;
Green = increase in satisfaction/dissatisfaction.
Satisfaction with the most recent housing assessment
in their area was highest among those who report their
house as being completely damaged (93%), followed
by those who say their house is not damaged (76%),
badly damaged (63%), and those with habitable houses
(56%). Satisfaction grew 15 points since IRM-2 among
46 Both have increased because the number of people who do not
know or who refuse to answer has declined since IRM-2.
those who said their house is habitable and 9 points
among those who say it is badly damaged. Levels of
satisfaction among those who said their house was
completely destroyed or not damaged was similar to
what was reported before.
Dissatisfaction with the CBS assessment was
high as the procedures and criteria were
unclear and because many thought that all
those who were deemed eligible for earlier
government assistance should have also been
included in the CBS list.
39


Aid and Housing Reconstruction Cash Grants
The qualitative research revealed that local percep-
tions of what makes someone an earthquake victim,
exacerbated by a lack of public knowledge of assess-
ment criteria, led to a feeling of injustice among
many who were not included in the beneficiary list.
As a result, the cash grant agreement process was
obstructed and delayed in many areas. Most protests
were to demand that those ‘unfairly’ excluded should
also be added to the beneficiary lists and were resolved
after people were informed that they could file official
complaints and be included later on if found eligible.
Government housing reconstruction cash grants
Many felt that they had been wrongly excluded
from receiving a reconstruction cash grant
under the Rural Housing Reconstruction
Program (RHRP).
The government is providing a reconstruction grant,
currently planned at NPR 300,000, as incentive to
build back better and to help offset some of the costs
of reconstructing houses. At the time the IRM-3 survey
was conducted, the size of the grant was to be NPR
200,000 and hence questions in this section ask about
a NPR 200,000 grant.
Most people—but not all—who said their house was
classified as fully damaged in the most recent damage
assessment said they have been declared eligible for
grant. However, 15% of people who said their house
was classified as fully damaged said they were not de-
clared eligible (Table 3.7). Among them, 93% resided
in the severely hit or crisis hit districts where benefi-
ciaries started to sign cash grant agreements under the
RHRP. Twenty percent of those who said their house
was partially damaged said they have been declared
eligible for the grant. No-one who said their house was
not damaged said they are eligible for the program.47
Table 3.7: Eligibility for RHRP grant - by housing damage classification (IRM-3, weighted)
Yes 1 No 1 Don’t know
Fully damaged 80% 15% 5%
Housing classification in the most recent Partially damaged 20% 74% 6%
official damage assessment Normal/Not damaged 0% 98% 1%
Don’t know 4% 74% 21%
Forty-nine percent of respondents said they were
declared ineligible for the RHRP grant. These people
were asked whether they should have been eligible for
it. Sixty-two percent of people who said they were not
eligible agreed that this was correct. However, 28% of
those declared ineligible said they should be eligible.
Feelings of being miscategorized as ineligible were
particularly high in the severely hit districts where,
overall, 83% of those declared ineligible said they
should have been eligible. Twenty-two percent of those
were told they are ineligible in crisis hit districts, 37%
in hit with heavy losses districts, and 24% in the hit
district felt that they had been unfairly excluded.
Almost half (47%) of those who felt they were unfairly
excluded said that their house was officially classified
as completely destroyed. This suggests that the prob-
lem is not just people disagreeing with how their house
was classified. While some people may not have un-
derstood what classification their house received, the
findings do suggest that there is a problem in ensuring
that those whose house was classified as completely
destroyed are eligible for the RHRP and that they un-
derstand they are. Twenty-two percent of those who
said they have unfairly been declared ineligible said
that their house was declared partly damaged.
Those who received the first installment of the
reconstruction cash grant generally received
the full amount, with the exception ofGorkha.
By September 2016, the government had begun dis-
bursing the first tranche of the reconstruction grant
(NPR 50,000) into bank accounts opened specifi-
cally for the purpose in the name of those who were
declared eligible and who had signed agreements.48
Importantly, the government and the NRA defined
disbursement of the housing grant as being the point
47 Results are similar if we look at people’s own classification on the
damage to their house. Seventy-six percent of respondents who
say their house was completely damaged say they were declared
eligible (40% badly damaged, 4% habitable, 0% not damaged).
48 See The Asia Foundation and Democracy Resource Center Nepal
(2016). Nepal Government Distribution of Reconstruction Cash
Grants for Private Houses: IRM - Thematic Study (November
2016). Kathmandu and Bangkok: The Asia Foundation, Section
3, for a fuller discussion.
40


Aid and Recovery in Post-Earthquake Nepal
at which the money was put in eligible beneficiaries’
bank accounts rather than when beneficiaries were
able to withdraw money.
Of those who had been able to access money at the
time the survey was conducted in September, most
received the full amount of the first installment (NPR
50,000) apart from in Gorkha (average NPR 49,872).
This suggests that some beneficiaries were charged a
fee, against the NRA guidelines, in Gorkha.
Delays and obstacles in accessing the cash
grants were common.
Few had received the first installment at the time of the
IRM-3 research (September 2016). Only 8% of those
who were declared eligible for the grant received any
money. This was largely due to the fact that disburse-
ment had not yet been completed and there has been
much progress in disbursing the money since then.
However, people interviewed for the qualitative re-
search also commonly mentioned problems accessing
the money, even when it was in their account. This was
particularly problematic for people who unable to go
to the bank in person, especially those who were out
of the country. Some district officers said that they
had asked the NRA to ease the access of earthquake
victims who were unable to visit banks in person to
withdraw the reconstruction grant, but had not yet
received replies from the NRA at the time of research.
In Okhaldhunga, a bank manager said that several
branch managers collectively proposed to ease the
process of distribution to earthquake victims in the
district, and to distribute cash to the victims in the
village, but that the NRA was not cooperative.
Mistakes while entering beneficiary details in cash
grant agreements were the most common problem
preventing people from withdrawing the first install-
ment from bank accounts. To illustrate, twenty-three
out of 286 beneficiaries could not access bank accounts
in Ramechhap due to the spelling of their names in
the cash grant agreements they signed not matching
the spelling in their citizenship certificate, their bank
account, and other documentation. Similar problems
were also reported in other districts.
Few planned to use the first installment of the
cash grant for the intended purpose.
The grant is to encourage earthquake-resistant
construction. Future tranches of funds are meant
to be dependent on following NRA guidelines on
safe construction. Of those who were declared eli-
gible, however, only 44% said they planned to do so
(Figure 3.6). One-quarter said they planned to use the
grant to rebuild or retrofit their previous house.49 Ten
percent said they would use the funds to support their
livelihoods and 5% to pay off loans. It is important to
note, however, that as the cash grant program was
being rolled out, many areas were without technical
assistance and few people would have been informed
about the precise building requirements.
Figure 3.6: Plans for use of RHRP grant amongst those declared eligible (IRM-3, weighted)
NRA model not sure if new
house will be
NRA model
49 There is now a separate grant to support retrofitting. Since late
2016, after IRM-3 research was conducted, some of those eligible
for reconstruction cash grants can use their first installment of
the grant for retrofitting instead (those categorized in damage
grade category 3-minor repairs). Overall, they would receive a
smaller amount disbursed in two rather than three installments.
41


Aid and Housing Reconstruction Cash Grants
The proportions of eligible beneficiaries who
plan to follow NRA rules on technical stand-
ards varied massively across districts.
In Nuwakot, 92% of people said they plan to do so
(Table 3.8). But in every other district, with the ex-
ception of Gorkha, more people said they will use the
money for other things. Planned retrofitting was par-
ticularly high in Ramechhap (53%) and Kathmandu
(50%). Building a house not following the NRA guide-
lines was high in Sindhupalchowk (17%). Planned use
to pay off loans was high in Lamjung (22%), Gorkha
(12%), and Dhading (10%). Use for livelihoods was
very high in Dhading (30%). Unsurprisingly, large
proportions of people in the hit with heavy losses
districts and the hit district, as well as in Kathmandu
and Bhaktapur, did not know what they would use
the money for. Grant disbursal had not started in
these places and people may have therefore had little
information on if and when the program would begin
and what the rules for it would be.
Table 3.8: Plans for use of RHRP grant amongst those declared eligible -
by district impact and district (IRM-3, weighted)
(D
Rebuild/retrofit previous house Build new house using accepted NRA model Build new house not using NRA model/not sure if new house will be NRA model Pay off loans Livelihood support For other things Don’t know
Severely hit 21% 55% 5% 6% 9% 2% 5%
Dhading 25% 30% 10% 10% 30% 0% 5%
Gorkha 8% 58% 6% 12% 12% 3% 5%
Nuwakot 3% 92% 3% 3% 8% 0% 0%
Ramechhap 53% 37% 1% 0% 1% 1% 8%
Sindhupalchowk 14% 53% 17% 0% 6% 3% 8%
Crisis hit 46% 3% 1% 1% 12% 0% 25%
Bhaktapur 0% 0% 0% 0% 0% 0% 0%
Kathmandu 50% 0% 0% 0% 12% 0% 25%
Okhaldhunga 7% 40% 7% 7% 13% 0% 27%
Hit with heavy losses 7% 24% 0% 12% 9% 0% 54%
Lamjung 0% 44% 0% 22% 11% 0% 33%
Solukhumbu 14% 0% 0% 0% 7% 0% 79%
Hit 0% 0% 0% 0% 0% 0% 100%
Syangja 0% 0% 0% 0% 0% 0% 100%
All districts 25% 44% 4% 5% 10% 1% 11%
Limited technical assistance was provided at
the time of research. Where deployed engi-
neers were present, they were often inactive
and waiting for cash grants to be distributed
and rebuilding to begin.
The qualitative research shows some people were
unwilling to wait for the housing reconstruction cash
scheme, which involved technical assistance, to be
rolled out and have started rebuilding on their own.
This usually meant that they did not use earthquake-
safe measures. The coverage of technical assistance
remained low at the time of IRM-3. Some non-
government organizations were providing technical
assistance alongside the government deployed
engineers. The latter were deployed during the cash
grant agreement process to brief citizens about
building codes and government criteria for house
reconstruction. But they were largely inactive in
September 2016, saying they had little to do as people
were waiting to receive their first installment before
starting rebuilding.
42


Aid and Recovery in Post-Earthquake Nepal
Satisfaction with the reconstruction cash grant scheme
Overall, perceptions of the RHRP housing
reconstruction program were not favorable.
People were more satisfied with the agreement
process than with the assessment to determine
eligibility or with access to the grant.
In each of the 24 wards in the six districts visited for
the qualitative research where the cash grant program
had been launched, researchers sought to determine
views towards the program, ranking each ward by
whether most people felt the program had been
satisfactory or not. They also sought to unpack views
towards different stages of the program.50
As Figure 3.7 shows, people were more likely to express
dissatisfaction with the program than satisfaction. In
15 wards, dissatisfaction was common compared
to only four where people were generally satisfied.
Dissatisfaction was particularly high about the CBS
damage assessment, which determined eligibility,
and about access to cash grants once people had
signed agreements. In contrast, people viewed the
cash grant agreement process more favorably. Many
thought the agreement process was well managed,
coordinated, and completed without any major
problems, after initial protests and obstructions were
resolved. The majority of citizens interviewed for
the qualitative research spoke positively of the cash
grant agreement process, including those not listed
as beneficiaries. People were satisfied with the process
of signing agreements at the VDC center, covering
wards one by one, and with the involvement of local
stakeholders such as local party representatives,
Ward Citizen Forum (WCF) coordinators, Social
Mobilizers, and other facilitators as well as I/NGOs
which provided substantial logistical support during
the process in most areas. People were also satisfied
with amendments to the NRA guidelines for signing
agreements such as new provisions for those out of
the country or without land ownership certificates.
Those who received the first installment of the grant
also assessed the program more favorably, although
they remained dissatisfied with the amount and
unclear information on timelines and procedures of
the different steps in the program.
Figure 3.7: Citizens’ perception of the government’s cash grant program
(number of wards - qualitative research)
Satisfied
Neutral H Dissatisfied
Frustration, discontent, and confusion was
particularly high among citizens and officials
in districts where the CBS assessment was not
conducted
Without completion of the CBS assessment, the cash
grant scheme could not yet be rolled out at the time
of research further delaying reconstruction in less
severely hit districts (Syangja and Solukhumbu visited
for the qualitative research). Citizens and officials in
Solukhumbu and Syangja criticized and expressed
frustration over the government’s decision to exclude
these districts from first round of the CBS assessment
delaying the cash grant agreement process, especially
50 The data is based on researchers’ conversations (individual
meetings, citizen interviews and focus group discussions) with
ward citizens, officials, and other stakeholders. As such it is
not necessarily an objective reflection on the quality of these
processes.
43


Aid and Housing Reconstruction Cash Grants
in Solukhumbu where at least seven VDCs registered
damages as high as those in severely hit districts.
Officials in both in Syangja and Solukhumbu were
frustrated with a lack of clarity and the government’s
inability to inform them if and when cash grants will
be provided to earthquake-affected households in
these districts.
Dissatisfaction was highest over the size
of the cash grant. While the grant was intend-
ed as incentive to build earthquake-resilient
buildings, not to fully cover construction
costs, many were dissatisfied with the amount
as they thought it was insufficient. Estimates
of construction costs show that the grant
will likely only cover a small share of the
costs.
Most said the amount of the reconstruction cash
grant was inadequate for covering construction costs
in any of the earthquake-affected areas. When those
who have been declared eligible for the grant were
asked to estimate costs for rebuilding/constructing,
the average amounts stated went well above the NPR
300,000 grant.51 The lowest average amount was NPR
404,019 in the hit district of Syangja. The average
cost mentioned in the severely hit, crisis hit, and hit
with heavy losses districts were NPR 1,014,626, NPR
2,523,949, and NPR 656,539, respectively.
The average cost of rebuilding/constructing people’s
house given by those whose house was classified as
fully damaged in the most recent damage assessment
was NPR 1,433,489. For those whose house was
classified as partially damaged, the figure was NPR
890,216. For those whose house was not extensively
damaged, but who were declared eligible for the grant,
the figure was NPR 280,632.
Seven in 10 respondents who were declared eligible for
the grant said that the NPR 200,000 grant would cover
less than one-quarter of the cost of reconstructing/
rebuilding their house. Two in 10 said it would cover
25-50% of the costs. Only 5% said this amount would
cover over half to all of the costs.
The majority of citizens interviewed in the qualitative
research were dissatisfied that the first installment
of NPR 50,000 would not be enough for even the
initial preparation for construction (demolition,
clearing debris, damp proofing coursing) let alone
constructing the foundations as prescribed in the
grant guidelines. Respondents in Sindhupalchowk and
Gorkha, in particular, said out that NPR 50,000 was
barely enough to demolish their old house. Concerns
regarding high carriage charges for transporting
building materials to remote areas and high labor costs
were also commonly raised.
The government has made provisions to pro-
vide soft loans to help with housing reconstruc-
tion but this has not happened in practice.
These loans can be of up to NPR 300,000 without
collateral. There is also provision for subsidized loans
of up to NPR 1,500,000 outside the Kathmandu
Valley, and up to NPR 2,500,000 inside the Valley
with collateral.52 However, banks have been found to
be reluctant to provide soft loans without assurances
from the government for repayment.53
Photo: Alok Pokharel
51 As noted, government policy has changed since the IRM-3 survey
with the grant now planned to be NPR 300,000. However,
questions were asked about the initially envisioned sum of NPR
200,000.
52 http://hrrpnepal.org/media/iO5469/i6ino_briefing-pack_v3.pdf
53 The Asia Foundation and Democracy Resource Center Nepal
(2016). Nepal Government Distribution of Earthquake
Reconstruction Cash Grants for Private Houses - Independent
Impacts and Recovery Monitoring Thematic Study. Kathmandu
and Bangkok: The Asia Foundation.
44


Aid and Recovery in Post-Earthquake Nepal
Case Study 3.1: Confusion over responding to complaints
in Okhaldhunga
In Okhaldhunga, informants said that conflict-
ing jurisdictions between different agencies
involved in reconstruction has added to the
delays in the reconstruction process. Govern-
ment officials and political parties involved
in reconstruction, who worked through the
DDRC prior to CBS assessment, blamed the
NRA for disregarding earlier assessments and
complaints. The DDRC had categorized 15,619
households as being fully damaged meaning
they would receive NPR 15,000 and NPR
10,000 cash assistance for temporary shelter
construction and for winter relief. Some 6,000
complaints against the categorization were
received by the DDRC. The CBS categorization
increased the list of beneficiaries to 19,818. Yet
even after the new assessment, around 5,600
additional complaints were filed. The district
level authorities said they were confused wheth-
er to address the complaints received by the
DDRC or the new complaints received after the
CBS categorization. They thought that the NRA
guidelines were confusing and officials have
struggled to address even genuine complaints
due to a lack of clarity.
Further, according district offices, almost 70
percent of the beneficiary details were wrongly
entered in the list. These mistakes were correct-
ed by VDC secretaries with computer operators.
A NRA sub-divisional representative appoint-
ed as a focal person to look into the issue in
Okhaldhunga could not continue his work due
to dissatisfaction over the NRA’s role and over
the exclusion of district level political parties
in the District Coordination Committee (DCC).
3.4 Coordination of aid
Coordination was generally weak at the
local level, both between different govern-
ment offices and between government and
non-governmental organizations. Overlap of
or confusion over respective responsibilities
hindered effective coordination and affected
the reconstruction process.
Many local government officials and other stakeholders
argued that the establishment of the NRA had hindered
coordination and the process of reconstruction. They
thought there was a lack of efficient coordination
between the NRA and local government offices and
that reconstruction could have been implemented
more efficiently through already established and
functioning government offices or mechanisms, such
as the District Disaster Relief Committees (DDRCs)
and Relief Distribution Committees. Delays in the
establishment of local NRA offices only made local
coordination more difficult. In several districts, local
NRA representatives were appointed only after the
completion of the CBS assessment and when the cash
grant agreement process had already started.54 At the
time of the research, some NRA sub-regional offices
also looked after neighboring districts. This too meant
that there was no local NRA presence in some districts,
hindering effective coordination and information
sharing there.
Overlapping jurisdictions and confusion over which
government office was responsible for what were
commonly observed. District and local level authorities
and political party representatives said that the fact
that NRA policies changed often and sometimes were
impractical or “unrealistic” made implementation
at the local level difficult. In their opinion this was
further compounded by their limited representation
or decision-making power in NRA platforms and the
54 For instance, the cash grant agreement process began in the first
week of May 2016 in Sindhupalchowk while the focal person of
NRA was appointed after mid-June. District level informants
in Ramechhap thought that the late opening of regional NRA
offices delayed the distribution of reconstruction cash grants.
No NRA representatives had been appointed in Syangja and
Solukhumbu districts at the time of research. Even though the
cash grant process had not yet started in the latter two districts,
the absence of a local NRA presence was seen to hinder efficient
communication on the process and the delivery of much-needed
information on timelines and procedures to local communities.
45


Aid and Housing Reconstruction Cash Grants
newly established Disaster Coordination Committees
(DCCs).55 * 55 The impact of this was particularly noticeable
with regards to complaints resolution (Case Study 3.1).
Filed complaints were often passed from one office to
another and back without being resolved.
The role of District Coordination Committees (DCCs),
established to facilitate the coordination of reconstruc-
tion at the district level, was generally ceremonial.
This body was found to be largely ineffective due to
the absence of members of parliament, who are gen-
erally based in Kathmandu and who are directed to
lead them. Local political party representatives whose
support local government officials tend to need to im-
plement decisions, were not invited to DCC meetings.
District government officials were often taking care
of the day-to-day work of the DCCs in the absence of
parliamentarians. Yet, they had little decision-making
power and ability to coordinate reconstruction in the
district through the DCCs.
Coordination between non-government organizations
or other actors involved in reconstruction and local
government bodies seemed to be poor in the six
districts visited for the qualitative research, despite
some attempts to make coordination smooth. I/NGOs
were widely criticized for not coordinating with district
or VDC stakeholders (government and civil society/
communities). In some cases, I/NGOs coordinated
with the district level authority but did not contact
VDC offices and local communities. This, government
officials and civil society representatives claimed, was
leading to a mismatch between I/NGO support and
people’s needs and increased dissatisfaction the 1/
NGOs. On the other hand, I/NGOs pointed out that
lengthy procedures to get their projects approved
often made timely and efficient coordination at the
local level difficult.
3.5 Communication and satisfaction
Communication about aid
The most common source for information
about aid were neighbors, radio, the VDC
office, and Ward Citizen Forums.
Of these, neighbors were a source of information
for 82% (Figure 3.8).56 Other top sources were radio
(31%), the VDC Secretary (24%), and Ward Citizen
Forum (WCF) members (18%). Political parties,
school teachers, and relatives and friends in district
headquarters or Kathmandu were less common
sources. Very few people got information on aid from
NGOs.
VDC offices and Ward Citizen Forums (WCFs) were
more important sources of information in more remote
areas. Fewer people in less remote areas relied on the
VDC Secretary (14%) and the WCF (11%) compared to
people in remote (31% VDC Secretary, 23% WCF) and
more remote (37% VDC Secretary, 29% WCF) areas.
The likelihood of political parties being a source of
information on aid also increases with remoteness
(4% less remote, 8% remote, 12% more remote areas).
Overall, people did not feel that they could
communicate well with aid providers, espe-
cially those removed from the local level.
When asked whether they felt they could communicate
to receive information or make a complaint, for every
aid provider people tended to say communication
was bad or, at best, okay. Relatively few said that
communication was good.
People were more likely to say communication was bad
with bodies that are the most removed from the local
level (Figure 3.9). Six in 10 said that communication
was bad with INGOs and foreign government (63%
each), and half of respondents said this about the
central government (50%). For other aid providers,
people tended to think that communication with them
was okay. Though few said that communication was
good, people were more likely to say this about the
police (29%), local administration centers57 (26%),
and the armed police force (24%).
55 New NRA platforms such as sub-regional NRA offices, DCCs, and
grievance hearing mechanisms do not formally involve local level
political parties. The DCCs are led by parliamentarians from the
respective districts. Local government officials were often leading
the day to day work of the DCCs in the absence of parliamentarians
who are based in Kathmandu. Yet, they had little decision-making
power and ability to coordinate reconstruction in the district
through the DCCs.
56 Multiple responses are allowed. Hence percentages do not add
up to 100%.
57 Refers to VDC office, ward level office in case of municipalities,
and area offices.
46


Aid and Recovery in Post-Earthquake Nepal
Figure 3.8: Sources of information on aid (IRM-3, weighted)
Figure 3.9: Satisfaction with communication with aid providers (IRM-3, weighted)
Religious groups
Foreign government
Private business
Local community
organizations
NGOs
INGOs
Local administration center
Local political parties
Armed police force
Police
Army
Central government
1 11% 31% 44% 13% 1%
| 7% 14% 63% 15% 1%
16% 35% 41% 7% 1%
22% 44% 29% 5%
13% 33% 47% 6% 1%
8% 18% 63% 10% 1%
26% 41% 30% 2%1%
22% 36% 38% 4%
24% 36% 35% 5%
29% 39% 29% 3%
22% 36% 37% 5%
18% 27% 50% 5%
I I I I I I
20%
40%
6o% 8o% ioo%
o%
Yes, communication is good H No, communication is bad Refused
Communication is okay Don't know
47


Aid and Housing Reconstruction Cash Grants
People in severely and crisis hit districts were more
likely to say that communication was good than
those in the less affected hit with heavy losses and
hit districts. In the severely hit districts, the aid
providers for which people were most likely to say
that communication was good are local community
organizations (27%), local administration centers
(26%), police (25%), and local political parties (24%).
In general, those who have not received aid were
slightly more likely to feel that aid communication
with most agencies was better than those who had
not received aid. This suggests that expectations on
communication are higher amongst aid recipients
than others.
Levels of satisfaction with communication
with aid providers were low.
Less than half of respondents were satisfied with how
any aid provider had informed them about aid since
the end of the winter (Figure 3.10). Respondents
were most likely to be satisfied with the police (51%),
followed by local community organizations (49%),
the army (47%), and the armed police force (46%).
People expressed the highest levels of dissatisfaction
on information provided about aid by political parties
(68%), private business groups (51%), and the central
government (50%).
People in the severely hit districts were more likely
than those in districts with lower levels of impact to
be satisfied with how aid providers have provided
information about aid. This points to aid providers
being more active in the severely hit districts compared
to other areas.
Satisfaction levels were higher if communication with
the particular aid provider was perceived as being
either good or okay (Table 3.9). This was especially
true for providers working in close proximity to aid
recipients, such as local administration centers and
local community organizations. For these bodies,
satisfaction was clearly tied to the perceived quality
of communication with half or more satisfied with aid
providers if communication is either okay or good.
This trend holds for local political parties, but the
level of satisfaction with them was very low regardless
of perceptions of communication with them. In
contrast, satisfaction with information providers did
not appear to be as linked to the perceived quality of
communication for providers more removed from the
area such as the central government, INGOs, NGOs,
and foreign governments.
Figure 3.10: Satisfaction with how aid providers communicate about aid -
by aid provider (IRM-3, weighted)
Religious groups
Foreign governments
NGOs
INGOs
Local administration center
Political parties
Police
Nepal Army
Central government
Private business groups
Local community
organizations
Refused
Don't know
I Somewhat unsatisfied
Very unsatisfied
Very satisfied
Somewhat satisfied
Armed Police Force
3% 31% 23% 15% 2% 26%
30% 29% 21% 1% 18%
r5% 44% 22% 15% 1% 13%
2% 39% 26% 16% 1% 15%
[3% 37% 27% 16% 1% 16%
14% 39% 28% 17% 1% 11%
|2% 20% 25% 43% 1% 10%
43% 23% 13% 1% 17%
| 4% 47% 22% 13% 1% 13%
44% 24% 13% 1% 15%
| 3% 38% 26% 24% 0% 9%
L2 1 1 1
100%
48


Aid and Recovery in Post-Earthquake Nepal
Table 3.9: Share of people satisfied with aid providers on how information on aid was given -
by whether people think communication was good or bad (IRM-3, weighted)
Yes, communication is good Communication is okay No, communication is bad
Central government 45% 51% 35%
Army 51% 59% 39%
Police 54% 60% 39%
Armed police force 49% 58% 38%
Local political parties 27% 28% 12%
Local administration center 53% 51% 27%
INGOs 37% 58% 35%
NGOs 41% 63% 31%
Local community organizations 53% 62% 31%
Private businesses 34% 48% 19%
Foreign governments 39% 63% 30%
Religious groups 35% 46% 19%
Different groups’ sources of inf ormation on
aid and satisfaction with these varied.
Those belonging to lower castes, those with lower
incomes, and those in remote areas were less likely to
say that they can communicate well with different aid
providers, to receive information or make a complaint.
Women were also slightly less likely than men to
say that communication was okay with various aid
providers asked about in the survey.
Satisfaction with aid distribution
Satisfaction with every aid provider decreased
significantly between IRM-2 and IRM-3.
Between IRM-1 and IRM-2 satisfaction levels with
most aid providers did not change dramatically
(Table 3.10). In February 2016 (IRM-2), eight in 10
respondents were satisfied with the security forces
(the army, police, armed police force), which was only
a slight decline from the high levels of satisfaction
with these bodies right after the earthquake during
rescue efforts. Satisfaction with local administration
centers nearly doubled between IRM-1 and IRM-2.
However, from March 2016 to September 2016 (IRM-
3), satisfaction with every aid provider decreased
sharply. Satisfaction levels with every aid provider was
below 50% in IRM-3, with the exception of the police
(51%). Satisfaction with other security forces, INGOs,
and NGOs dropped by at least 30 percentage points
between IRM-2 and IRM-3. The smallest change in
satisfaction was with political parties (5 point drop),
but the level of satisfaction with them was already
comparatively low with one-quarter being satisfied
with political parties in IRM-2.
In September 2016, the government, INGOs, and
NGOs were mentioned as the most common aid pro-
viders. However, only four in 10 people were satisfied
with any of these bodies. The level of satisfaction with
them was similar to that with foreign governments and
lower than any of the security forces or local commu-
nity organizations, all entities that provided much less
aid than the government, INGOs, or NGOs.
Table 3.10: Proportion satisfied with aid provider
(IRM-1, IRM-2, IRM-3, weighted)
I IRM-1 I I IRM-2 I I IRM-3 I
Central government 56% 51% 40%
Army 90% 83% 48%
Police 90% 82% 51%
Armed police force 88% 80% 47%
Local political parties 36% 26% 21%
Local administration center 33% 60% 43%
INGOs 75% 73% 39%
NGOs 69% 70% 41%
Local community organizations 63% 66% 49%
Private businesses 53% 51% 29%
Foreign governments 72% 67% 40%
Religious groups 51% 53% 26%
Satisfaction with the central government, INGOs,
and NGOs was higher among those who received aid
in IRM-3. These bodies were also the top providers
of aid, which likely drove the favorable views. Other
providers got mixed reviews. Satisfaction with the
security forces and foreign governments was higher
among those who did not get aid. Levels of satisfaction
49


Aid and Housing Reconstruction Cash Grants
Photo: Chiran Manandhar
were similar among those who got aid and those
who did not when it comes to assistance provided by
political parties, local administration centers, local
community organizations, private businesses, and
religious groups.
People said dissatisfaction with I/NGOs was
rising because of their alleged disregard of
people’s needs when designing and imple-
menting programs.
Dissatisfaction with non-governmental recovery
programs was mainly due to the high priority people
gave to the reconstruction of private houses. Where
I/NGOs worked on reconstruction, it was largely
focused on relief or trainings and other soft forms of
assistance rather than directly rebuilding or providing
cash grants for rebuilding houses. People did not
always think such assistance was needed or useful.
As a local resident in Sindhupalchowk said: “Life
does not move ahead with bucket and soap. The time
to distribute such materials is over. Organizations
should start distributing construction materials if
they really want to help the earthquake victims.”
Relief based on specific procedures and criteria set
by I/NGOs for targeting also drew criticism from
those excluded. Dissatisfaction was higher and more
strongly expressed in Gorkha and Sindhupalchowk
where I/NGO presence was higher.
Increasing dissatisfaction with the govern-
ment and political parties was largely due
to delays in the provision ofcash grants, un-
clearpolicies and information, and delays in
addressing complaints.
50


Aid and Recovery in Post-Earthquake Nepal
In Syangja and Solukhumbu, where the cash grant
agreement process had not yet begun, dissatisfaction
was highest. People in these districts were highly
dissatisfied with the government for delays in rolling
out and information on the reconstruction cash
grant program. Even in districts where cash grants
were being distributed, people were dissatisfied with
the government. This was mainly due to what they
considered to be flaws in the CBS assessment, being
missed out of the beneficiary lists, delayed cash grant
distribution, what they considered to be insufficient
cash support, and the delayed or unclear process of
addressing complaints. At the time of the research, few
complaints had been resolved and most were passed
to the next higher office or the NRA. Most of those
who had filed grievances had not heard anything on
whether and how their grievances would be resolved.
Much progress has since been made in addressing
complaints. Yet many still need further verification
or re-assessment.
the largest drop was in Sindhupalchowk with just four
in 10 agreeing with the statement in IRM-3 compared
to over seven in 10 in IRM-1 and IRM-2. There was
also a sharp drop in Dhading and Ramechhap while
views were similar to IRM-2 in Gorkha and Nuwakot.
In the crisis hit districts, views among respondents
surveyed in all three waves of the survey in Bhaktapur
and Okhaldhunga remained unchanged, but there was
a steep drop in the share believing aid distribution
has been fair in Kathmandu. Kathmandu has by the
far the lowest level of satisfaction of any district. The
perception of people thinking distribution was fair
increased between IRM-1 and IRM-2 in Solukhumbu
and Syangja, but dropped in IRM-3 in both districts.
Fewer people than before thought that every-
one could get aid according to their needs
than in the past but those in severely hit dis-
tricts were more likely to agree that everyone
could get aid according to their needs.
Table 3.11: Share of people who agree
that VDC/municipalities are distributing
aid fairly - by district impact and district
(IRM-1, IRM-2, IRM-3 household panel, unweighted)
I IRM-1 I I IRM-2 I I IRM-3 I
Severely hit 72% 73% 57%
Dhading 69% 69% 50%
Gorkha 81% 72% 68%
Nuwakot 64% 81% 78%
Ramechhap 73% 74% 57%
Sindhupalchowk 73% 74% 40%
Crisis hit 46% 52% 50%
Bhaktapur 36% 42% 45%
Kathmandu 28% 53% 19%
Okhaldhunga 63% 63% 63%
Hit with heavy losses 58% 61% 49%
Lamjung 55% 54% 45%
Solukhumbu 67% 89% 67%
Hit 51% 72% 45%
Syangja 51% 72% 45%
All districts 63% 67% 54%
Perceptions of the fairness of the distribution
of aid by VDCs or municipalities also markedly
declined.
Dissatisfaction with VDC offices was also high. Most
often, residents complained about the absence of
VDC secretaries. There were also complaints about
the accountability of the VDC office and a lack of
information. In IRM-1, 63% of the 1,470 people who
were also interviewed in IRM-2 and IRM-3 believed
distribution was fair and this increased to 67% in IRM-
2. However, this then declined to 54% by the time of
IRM-3 (Table 3.11). Among the severely hit districts,
Of the 4,446 respondents interviewed in both IRM-2
and IRM-3,75% agreed (26% strongly, 49% somewhat)
and 20% disagreed (2% strongly, 18% somewhat) with
the statement that people of every caste, religion, and
ethnicity were equally able to receive aid according
to their needs in IRM-3. The share agreeing with the
statement decreased from 90% in IRM-2 to 75% in
IRM-3 points (Figure 3.11).
Figure 3.11: Opinions on whether everyone
can get aid according to their needs
(IRM-2, IRM-3 household panel, unweighted)58
100% 1-
80% -
60% -
40% -
20% -
1
Strongly agree
Somewhat agree
j
I Strongly disagree
Somewhat disagree
58 Bars do not add up to 100% because some respondents did not
have an opinion.
51


Aid and Housing Reconstruction Cash Grants
Marginalized and vulnerable groups were
frequently identified as groups less likely to
receive aid according to their needs.
Those who disagree that everyone is able to get aid
equally according to their needs were asked who is
less likely to receive aid according to their needs.
People most commonly mentioned a caste group:
lower caste (45%), higher caste (16%), and Janajatis
(17%) — Figure 3.12. Other groups named include
the elderly (27%), women (14%), and those who are
disabled/sick (14%).
Figure 3.12: Groups who are unable to get aid equally according to their needs among those
who disagree that everyone can get aid equally (IRM-3, weighted)
52


Photo: Alok Pokharel

4.1 Borrowing
Rates of borrowing
The proportion of people borrowing in IRM-
3 remained similar to IRM-2 but was much
higher than in IRM-1.
The number of people borrowing in IRM-3 remained
high and was similar to IRM-2 at 32%,59 a doubling of
the numbers since IRM-1. Almost half of the popula-
tion in severely hit districts had borrowed in IRM-3.
Around one-quarter reported borrowing in crisis hit
and hit with heavy losses districts.
Amounts borrowed have increased since the
early months after the earthquakes.
Over half of those who took loans between IRM-2
and IRM-3 borrowed less than NPR 100,000 (59%).
Twelve percent borrowed between NRP 100,000 and
200,000 and seven percent borrowed between NRP
200,000 and 400,000. Only ten percent borrowed
more than NPR 400,000.
Weighted data show that the average amount borrowed
decreased from NPR 303,130 in IRM-2 to NPR 213,451.
However, this still far exceeds the average amounts
borrowed in IRM-1 (NPR 103,057) - Table 4.2. The
overall decline in sums borrowed between IRM-1 and
IRM-3 was driven mainly by Dhading, Kathmandu,
and Solukhumbu. Elsewhere, the average amount
borrowed increased.
Table 4.1: Share of people who borrowed money
(IRM-1, IRM-2, IRM-3, weighted)
1 IRM-1 I I IRM-2 I I IRM-3 I
Severely hit 24% 49% 43%
Dhading 25% 52% 48%
Gorkha 17% 45% 36%
Nuwakot 14% 43% 34%
Ramechhap 40% 63% 59%
Sindhupalchowk 30% 46% 42%
Crisis hit 11% 22% 25%
Bhaktapur 11% 22% 14%
Kathmandu 9% 19% 23%
Okhaldhunga 30% 66% 66%
Hit with heavy losses 10% 24% 24%
Lamjung 7% 21% 23%
Solukhumbu 15% 29% 26%
Hit 4% 43% 45%
Syangja 4% 43% 45%
All districts 14% 32% 32%
59 IRM-3 borrowers took loans between the end of the 2016 winter
season (around March 2016) and September 2016. IRM-2
borrowers took loans from the beginning of the 2015 monsoon
season (June 2015) and March 2016. IRM-1 borrowers took loans
between the April earthquake and June 2015.
53


Aid and Housing Reconstruction Cash Grants
Table 4.2: Average borrowing in NPR - by district
impact and district (IRM-1, IRM-2, IRM-3, weighted)
1 IRM-1 I I IRM-2 I I IRM-3 I
Severely hit 45,289 262,343 155,094
Dhading 54,719 645,171 172,533
Gorkha 53,910 149,389 152,641
Nuwakot 38,668 153,974 176,446
Ramechhap 44,811 118,267 121,906
Sindhupalchowk 34,859 111,245 150,104
Crisis hit 185,747 408,363 300,829
Bhaktapur 66,671 213,744 573,812
Kathmandu 243,843 531,259 324,193
Okhaldhunga 49,740 97,622 110,859
Hit with heavy losses 99,799 186,422 216,281
Lamjung 62,071 228,662 305,088
Solukhumbu 130,514 131,100 75,000
Hit 34,375 167,021 194,430
Syangja 34,375 167,021 194,430
All districts 103,057 303,130 213,451
Unweighted panel data from the three research rounds
show that the sums borrowed from different sources
have been increasing. Table 4.3 outlines the average
amounts borrowed, disaggregated by earthquake
impact category. It shows that, in general, people
have been borrowing more from most sources across
each category of earthquake impacts. For instance,
the mean amount borrowed from banks in severely
hit districts increased almost ten-fold from IRM-1 to
IRM-3. Increases for other lending sources were not
as steady, but there is hardly any category where the
mean average declined compared to IRM-1.
Table 4.3: Mean of self-reported amount (in thousand NPR) borrowed from different sources
in the three survey waves (IRM-1, IRM-2, IRM-3 household panel, unweighted)60
| IRM-1 | IRM-2 | | IRM-3 | | IRM-1 | | IRM-2 | IRM-3 | | IRM-1 | | IRM-2 | | IRM-3 IRM-1 IRM-2 IRM-3
Severely hit Severely hit Severely hit Crisis hit Crisis hit Crisis hit Hit with heavy losses Hit with heavy losses Hit with heavy losses z z z
Moneylender 65 94 103 34 228 83 11 28 48 20 200
Friend 73 68 75 90 65 1348 175 133 30
Relative 44 102 144 98 78 100 20 121 183 28 150 55
Neighbor 34 151 92 14 70 81 20 109 33 28 153 172
Other individuals 25 104 72 14 25 1500
Bank 56 301 565 464 303 400 520 1286 50 177 109
Savings and credit group 37 63 108 32 85 95 80 158 81 10 71 59
Co-operatives 84 86 78 48 161 302 48 53 120 5 20 216
Other financial institution 13 69 59 115 58 20 29
Sources of credit
Cooperatives, savings and credit groups, and
neighbors were the most common sources
of credit. Average monthly interest rates
for many sources, especially informal ones,
increased slightly since IRM-2 suggesting a
growing demandfor credit.
and neighbors (19%) - Figure 4.1. The decline in the
share of lending by friends and relatives observed in
IRM-2 continued, with 13% of borrowers taking loans
from their relatives in IRM-3. The share of borrowers
The most common sources of credit in IRM-3 were -----------------------
cooperatives (23%), savings and credit groups (20%), 60 Blank cells mean no-one borrowed from this source.
54


Aid and Recovery in Post-Earthquake Nepal
who took loans from banks stayed the same as in IRM-
2 (13%) while slightly more people were taking loans
from moneylenders (12% compared to 10% in IRM-2).
Banks are reportedly often reluctant to provide loans
when people approach them. Some pointed out that
banks find it challenging to collect regular installments
from clients in rural villages and do not trust them to
pay regularly, hence they only provide loans to bigger
entrepreneurs. Banks were perceived to be particularly
inaccessible to displaced communities such as Dalit
households who do not have any collateral.
Figure 4.1: Sources of borrowing among those who borrowed (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
Table 4.4: Average borrowing in NPR - by sources
(IRM-1, IRM-2, IRM-3, weighted)
IRM-1 mean IRM-2 mean IRM-3 mean
Moneylender 66,009 763,730 107,966
Friend 55,080 99,064 462,343
Relative 156,562 102,836 208,144
Neighbor 123,576 103,889 103,631
Other individual in ward 24,534 97,546 154,018
Bank 87,196 887,654 488,050
Savings and Credit group 53,888 109,503 98,616
Co-operatives 65,396 161,435 212,858
Other financial institution 11,522 130,528 48,458
Government loan scheme - 12,696 -
Average monthly interest rates have increased slightly
since the earthquakes. Interest rates for banks
remained relatively stable and low, but have slightly
increased from 1.5% in IRM-1 to 1.52% in IRM-2 to
1.73% in IRM-3. Interest rates from informal sources
like moneylenders, friends, relatives, neighbors, and
other individuals also increased since IRM-1. This
suggests a growing demand for credit from these
sources over time. Mean interest rates for savings and
credit and other financial institutions were relatively
higher than other sources in IRM-1 but have declined
gradually from 1.88% and 2.19% in IRM-1 to 1.63%
and 1.64% in IRM-3.
The average amounts borrowed from money
lenders, banks, saving and credit groups,
and other financial institutions significantly
decreased, while there was an increase in the
amount borrowed from friends, relatives,
other individuals, and cooperatives.
The average amount borrowed from each lender
changed significantly between IRM-2 and IRM-3
(Table 4.4). The average sum borrowed from mon-
eylenders and banks declined significantly between
IRM-2 and IRM-3, although banks still lend the high-
est amount to each borrower on average. In contrast,
sums borrowed from friends increased four-fold and
the average amount borrowed from relatives doubled.
Lowest average borrowing in IRM-3 was from savings
and credit and other financial institutions. It is likely
that higher interest rates charged by these financial
institutions compared to banks in IRM-1 and IRM-2
may have attracted people to banks where they can
get loans from them.
55


Aid and Housing Reconstruction Cash Grants
Some expressed apreferencefor informal
sources of credit despite higher interest
rates as these were seen to be more easily
accessible.
The qualitative research revealed that in most VDCs,
people found borrowing from family, friends, neigh-
bors, and individual moneylenders more convenient,
even though they charged higher interest rates than
banks. Borrowing locally from informal sources is
faster and easier than approaching banks, which are
generally located further away and require formal
documents and collateral. Further, villagers often
lack knowledge on how to approach formal sources
of credit and feared these lenders might take away
their land if they failed to pay back their loan in time.
“There are a couple of reasons why people resort to
moneylenders for loans. The most important one is
ease and convenience [...] moneylenders are readily
available in local areas. Secondly, for those who do
not have resources and property to keep as collateral,
the banks do not provide loans.” Informal sources
were also more flexible not having the strict payback
deadlines of banks and formal financial institutions.
“There is no risk of losing collateral if one is unable to
payback on time. Sometimes we can request the local
moneylenders to postpone or waive some interest and
there is no need for paperwork or collateral.”
Borrowing needs
Livelihoods, food, and rebuilding houses were
the main reasons for borrowing. Shelter-re-
lated borrowing (temporary shelter, rebuild-
ing houses, improving temporary shelters)
was concentrated in the severely hit districts.
Borrowing occurred for various purposes, both related
and unrelated to the earthquakes. Respondents
generally borrowed for household expenses and
income generation, particularly to open small
businesses and for labor migration abroad - a trend
continuing from before the earthquake. They also
borrowed because of financial stress incurred by
the earthquakes. Several took loans to reinvest in
businesses destroyed by the earthquakes, others to buy
livestock they had lost, or to build temporary shelters
and reconstruct their houses. Some had to borrow for
consumption as the earthquake initially affected their
income sources.
As with IRM-2, livelihoods, food, and rebuilding
houses were the main reasons for borrowing in IRM-
3. Of those who borrowed, 55% in IRM-2 and 58% in
IRM-3 said they borrowed to support their livelihoods,
the most common reason for taking loans in districts
in all earthquake impact categories (Figure 4.2).
Rebuilding houses was the second most common
reason for borrowing in IRM-2 but declined in IRM-3.
Twenty-six percent of people who borrowed in IRM-3
said they had borrowed to buy food. Only 11% of those
who borrowed mentioned financing temporary shelter
as the reason why they took a loan in IRM-3.
Figure 4.2: Reasons for borrowing, share of those borrowing (IRM-2, IRM-3, weighted)
56


Aid and Recovery in Post-Earthquake Nepal
While many were borrowing to deal with
earthquake impacts, the number of those
with heavily damaged houses who borrowed
decreased between IRM-2 and IRM-3. Aid
received by these households may explain
this decline.
Reconstruction was one of the main reasons why
people borrowed and the extent of damage to people’s
house from the earthquakes correlates with the likeli-
hood of borrowing, suggesting people were borrowing
to deal with the impacts of the quakes. In all three
surveys, people were more likely to borrow if they
experienced larger earthquake impacts (Figure 4.3).
However, while the more affected were still more
likely to borrow than others, there was a decrease in
the proportion of those whose house was heavily dam-
aged who borrowed after IRM-2. Aid received by these
households may explain this decline. Results show that
those who received aid between IRM-2 and IRM-3
were far less likely to have borrowed money in that
period (15%) than those who did not receive aid (87%).
Figure 4.3: Share of people who have borrowed - by housing damage (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
Borrowing is likely to increase over time.
Many said they planned to borrow in the
near future to cope with the impacts of the
earthquakes.
Findings from both research components suggest that
borrowing is likely to increase. Many said they would
need to borrow to cover costs for reconstruction,
especially if they did not receive the reconstruction cash
grant and if the government did not provide interest-
free loans for earthquake victims. Unsurprisingly,
many of those who were likely to borrow repeatedly—
people in severely hit districts, more remote and rural
areas, with low incomes and of low castes—were also
more likely to say they planned to borrow in the next
three months. Thirty-five percent of people in IRM-3
mentioned that they planned to borrow in the next
three months. People in the more affected severely
hit districts (60%) were much more likely to say they
planned to borrow.
People in more remote areas were more likely to say
they planned to borrow in the next three months. Fifty-
three percent of individuals in more remote regions
planned to borrow compared to 40% in remote and
26% in less remote areas. Similarly, 42% in rural areas
plan to borrow compared to only 22% in urban areas.
Other socio-economic factors similarly correlated
with borrowing intentions. Individuals who were
less educated and those of lower caste or lower
income were more likely to express a plan to borrow
money in the next three months. People who were
more educated expressed lower levels of borrowing
intentions, while a higher share of the less educated
wanted to borrow in the future. Similarly, 48% of
those in the low income group intended to borrow
while only 22% of those with a high pre-earthquake
income intended to do so. A larger share of people
with a disability (50%) mentioned a plan to borrow
more money in the next three months, compared to
only 35% people who have no disability.
57


Aid and Housing Reconstruction Cash Grants
Who was borrowing?
Borrowing was higher among already
vulnerable groups. People in remote and
rural areas, in severely hit districts, low
caste individuals, those with lower income,
and those in temporary shelters reported
higher rates of borrowing. These groups were
borrowing more frequently, at higher interest
rates, and were more likely to say that they
planned to borrow in the nearfuture.
While borrowing is generally common, some groups
had higher rates of borrowing. People in more remote
areas were more likely to borrow. As Figure 4.4 shows,
42% of people in more remote areas borrowed in
IRM-3 compared to 40% in remote areas and 21% in
less remote areas. This pattern was also observed in
IRM-2 but not in IRM-1. Remote regions in IRM-1
had the highest rate of borrowing (35%) compared
to less remote (13%) and more remote areas (21%).
However, in all three surveys, less remote regions
had lower borrowing compared to remote and more
remote regions. The pattern is also clear when looking
at differences in borrowing between rural and urban
areas across the three surveys (see Figure 4.4).
Figure 4.4: Share of people who have borrowed - by rural/urban
and remoteness (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-3 â–  IRM-2 â–  IRM-1
Figure 4.5: Share of people who have borrowed since the end of winter -
by where people are living (IRM-3, weighted)
own land people’s land public land
58


Aid and Recovery in Post-Earthquake Nepal
Poorer people, low caste individuals, and daily
wage laborers also reported particularly high rates
of borrowing in IRM-3. Those living in temporary
shelters were also more likely to borrow than others.
While 46% of low caste people borrowed between
IRM-2 and IRM-3, 30% of Janajatis and 33% of high
caste people reported the same. Those working as daily
wage laborers (45%) and in agriculture (39%) were
the most likely to have borrowed. Those who had no
job are also more likely than most to borrow (41%).
Those with a low income had the highest borrowing
rate (40%) compared to those with medium (32%) and
high (26%) income.
Table 4.5: Share of people who borrowed money -
by caste, occupation and income (IRM-3, weighted)
Proportion borrowing
High caste 33%
Caste Janajati 30%
Low caste 46%
Agriculture 39%
Industry/business 23%
Service 31%
Labor 45%
Occupation Student 15%
Housewife/ house-maker 25%
Retired 19%
Unemployed 41%
Low 40%
Income Medium 32%
High 26%
Current living conditions also had an impact on
borrowing patterns. Those living in temporary shelters
were more likely to borrow than others (Figure 4.5).
While 28% of those living in their own houses say
they have borrowed, more than 40% of those who are
living in shelters on their own or on other people’s
land mention having borrowed money.
Individuals in severely hit districts, remote and rural
areas, and those with lower income, were also more
likely to borrow repeatedly compared to those in other
districts, urban and less remote regions and those with
high incomes. When examining the household panel
dataset, which includes people interviewed in the last
two surveys (IRM-2 and IRM-3), there was significant
variation in terms of how frequently people borrowed.
Nearly 26% of people borrowed in both surveys, 34%
borrowed in one of the two surveys, and 40% did not
borrow in either. Okhaldhunga and Ramechhap were
the two districts with the highest shares of people
borrowing during both IRM-2 and IRM-3 (Table 4.6).
More people (31%) borrowed in both time periods in
more remote areas compared to only 28% in remote
and 16% in less remote districts. Similarly, 28% of
people in rural areas borrowed in both time periods
compared to only 7% respondents in urban areas. It
is most likely that this high demand is the reason why
people in more remote and rural areas report having to
pay higher interest rates. However, this also suggests
economic hardship faced by individuals living in
remote and rural regions where demand for capital is
induced by the natural disaster.
Table 4.6: Share of people who borrowed
in both IRM-2 and IRM-3 - by district
impact, district, rural/urban and remoteness
(IRM-2, IRM-3 household panel, unweighted)
Did not borrow Borrowed once Borrowed twice
Severely hit 34% 37% 29%
Dhading 28% 44% 28%
Gorkha 41% 36% 23%
Nuwakot 43% 37% 21%
Ramechhap 21% 37% 42%
Sindhupalchowk 39% 36% 25%
Crisis hit 42% 29% 29%
Bhaktapur 69% 25% 6%
Kathmandu 60% 32% 9%
Okhaldhunga 18% 30% 52%
Hit with heavy losses 58% 32% 9%
Lamjung 64% 26% 9%
Solukhumbu 51% 40% 10%
Hit 40% 32% 28%
Syangja 40% 32% 28%
All districts 40% 34% 26%
Less remote 53% 31% 16%
Remote 36% 36% 28%
More remote 36% 33% 31%
Rural areas 37% 35% 28%
Urban areas 66% 27% 7%
Thirty percent of people in the lower income group
reported borrowing in both surveys compared to 24%
in the medium income group and 20% in the higher
income group (Figure 4.6). In contrast, almost half
of the population in the high income group (47%)
reported not borrowing in both rounds, compared to
35% in the low income group. In short, individuals in
remote and rural areas, and those with lower income,
were more likely to borrow repeatedly compared to
those in urban and less remote regions and those with
high incomes.
59


Aid and Housing Reconstruction Cash Grants
Figure 4.6: Frequency of borrowing across the last two surveys - by income
(IRM-2, IRM-3 household panel, unweighted)
Low Medium H High
People in severely hit and hit districts, rural,
and more remote areas were more likely to
pay higher interest rates.
Interest rates charged were generally higher in rural
areas than in urban areas (Table 4.7). They were much
higher in more remote areas. Thirty-six percent of
borrowers from those places mentioned that average
interest rates are above 2%, compared to only 14% in
remote and 7% in less remote areas. In contrast, 22%
in less remote areas said that interest rates are less
than 1%, compared to only 12% in remote and 13% in
more remote areas. This distribution of interest rates
suggests that there is a higher need for capital in rural
and remote areas, where the market is relatively less
competitive compared to urban and less remote areas.
Some of the same groups that reported higher
rates of borrowing were also more likely to
face unsuccessful borrowing. People in se-
verely hit districts, those living in temporary
shelters on public land, and those of low in-
come and low caste groups were more likely
to be refused loans. Land, the most common
collateral, was used more frequently among
high income groups who tend to borrow larg-
er amounts of money.
Loan refusals have continued to remain low, only 3%
in IRM-3. People in severely hit districts were more
likely to have been unsuccessful than others: 4% were
refused loans, compared to only 2% from crisis hit
and hit with heavy losses districts and 1% from the
hit district. A higher share of people who do not live
in their own houses were unsuccessful borrowers
(Figure 4.7). Those living in shelters on public land
were particularly likely to have been unsuccessful in
their borrowing attempts.
Table 4.7: Mean reported interest rates - by
district impact, district, rural/urban and remoteness
(IRM-3, weighted)
Less than 1% Between 1%-1.5% Between 1.5%-2% £ CM O _Q < Refuse
Severely hit 13% 16% 46% 23% 2%
Dhading 12% 19% 58% 11% 0%
Gorkha 19% 13% 46% 20% 1%
Nuwakot 9% 13% 57% 21% 0%
Ramechhap 5% 18% 44% 33% 1%
Sindhupalchowk 19% 16% 26% 30% 8%
Crisis hit 20% 22% 31% 6% 21%
Bhaktapur 58% 30% 8% 0% 4%
Kathmandu 19% 24% 26% 4% 27%
Okhaldhunga 7% 8% 63% 22% 0%
Hit with heavy losses 15% 16% 28% 41% 0%
Lamjung 13% 22% 27% 38% 0%
Solukhumbu 17% 7% 30% 46% 0%
Hit 5% | 15% 76% 4% 0%
Syangja 5% 15% 76% 4% 0%
All districts 15% 19% 42% 15% 10%
Rural areas 16% 17% 43% 17% 6%
Urban areas 13% 23% 37% 6% 22%
Less remote 22% 21% 38% 7% 12%
Remote 12% 20% 43% 14% 10%
More remote 13% 8% 41% 36% 1%
60


Aid and Recovery in Post-Earthquake Nepal
Figure 4.7: Unsuccessful borrowers - by where people are living (IRM-3, weighted)
Lower income and low caste people were also more
likely to have been unsuccessful in borrowing. Five
percent of those with low incomes were unsuccessful
borrowers compared to only 2% of the medium income
group and 1% of those in the high income group.
Similarly, low caste people were twice as likely to
be unsuccessful borrowers as those from high caste
groups (2%) and 1% more than the Janajatis.
The two most stated reasons why individuals failed
to get credit61 were creditors refusing to grant credit
and the terms of credit being too hard to meet. Since
relatively larger shares of low income and low caste
people were unsuccessful borrowers, Table 4.8 pre-
sents the reasons disaggregated by income level and
caste. Forty-eight percent of low income people who
were unsuccessful in their borrowing attempts men-
tion that the creditor refused without specific reasons
and 39% said that the terms of credit were too hard
to meet. In contrast, only 10% of unsuccessful high
income borrowers faced refusal from creditors while
20% felt that the terms of credit were hard to meet.
Janajatis were more likely than others to mention the
two primary reasons for their failure to secure loans.
Table 4.8: Reasons for unsuccessful borrowing - by income and caste (IRM-3, weighted)
Creditor refused credit Terms of credit too hard to meet Process is too difficult Refused Don’t know
All unsuccessful borrowers 38% 37% 19% 1% 21%
Low income 48% 39% 12% 1% 13%
Medium income 34% 43% 38% 3% 16%
High income 10% 20% 4% 0% 65%
High caste 15% 27% 10% 0% 50%
Janajati 48% 44% 23% 2% 10%
Low caste 48% 24% 18% 4% 6%
The majority of those who borrowed (89%) did
not provide any collateral. But most of those who
borrowed (59%) were taking on loans of less than NPR
100,000. People borrowing larger amounts were more
likely to need collateral to secure loans. Most people
borrowing from banks in IRM-3 provided some form
of collateral for their loans. In contrast, more than
90% of individuals who borrowed from relatives,
neighbors, or other individuals have not provided
collateral. Amongst those who did provide collateral,
land was the most frequent form provided (8%). One
percent said they used their house as collateral and
the same proportion of people say they used jewelry
or household items. The use of land as collateral was
more frequent among borrowers with a high income
(12%) compared to those in the low income group
(6%). High caste people were also more likely to use
land as collateral (11%), compared to Janajatis (7%)
and low caste people (5%).
61 This was a multiple choice question. As such, percentages for
stated reasons sum to more than 100%.
61


Aid and Housing Reconstruction Cash Grants
Dalits faced particular difficulties accessing
credit especially from formal sources.
Dalits were found to be struggling more and recovering
more slowly than others; many had barely enough
money and food for daily needs. Despite being most in
need of credit, this group was the least likely to be able
to access loans from formal sources. Dalits are often
illiterate, lack collateral, and do not have access to
social networks and information required to approach
financial institutions for loans. Those who managed to
take loans mostly borrowed from local moneylenders
charging particularly high interest rates. However,
many had insufficient income to pay interest and
pay back their loans. As such, Dalits are particularly
vulnerable to debt traps.
Rising debts were a worry for many house-
holds and the risks of debt traps were in-
creasing.
As more were borrowing larger amounts and from
informal sources, often at high interest rates, debt
burdens were increasing for many earthquake-affected
households. According to findings from the qualitative
research many were unsure how they would be able
to pay back their loans. Households were found to be
struggling to repay loans. For example, an old man
in Syangja said, “We built this house after our house
completely fell down because of the earthquake. We
are still in debt. I don’t know when we will be able to
repay it.” Another man in the same district added,
“Now the house stands incomplete, I do not have any
source of income and on top of that we have to pay
interest to the bank. I am now worried about paying
back the loan.” The large amounts that people had to
borrow for reconstruction meant that loans could not
be paid off merely by increasing incomes from farm-
ing, labor, or small business as people commonly did
before the earthquakes.
Most households said relied on government assistance
in the form of cash grants and soft loans or remit-
tances to pay back loans. Across districts, households
often mentioned the need for remittances to pay off
larger debts incurred by reconstruction-related costs
(Case Study 4.1). “Without loans, houses cannot be
built and they cannot be paid back without going
abroad,” explained a man in Syaule VDC in Sindhu-
palchowk.
Case Study 4.1: A migrant laborer borrowing to rebuild
Yagya Bahadur Ale Magar from Prapcha in
Okhaldhunga has been working in Saudi Ara-
bia for five years. His two brothers also work
there. They were unable to return home after
the earthquakes destroyed his family’s home as
the company did not grant them leave. When
Yagya Bahadur’s mother, wife, children, and
sister-in-law were forced to stay in a tempo-
rary shelter, he wanted to send them more
money to improve their shelter and eventually
rebuild. Normally, he would send around NPR
20,000-30,000 every other month. However,
this was not enough to make ends meet after the
earthquakes and he and his brothers borrowed
money from other Nepali migrant laborers in
Saudi Arabia and sent around NPR 100,000 to
their family. He said, “[migrant workers from
the same village] also tried to collect money
but could not get enough, that’s why I sent as
much as I could.” He explained that, “The Saudi
government allowed earthquake victims to send
free money transfers, that’s why it was easy to
send money.”
Over one year after the earthquakes, Yagya
Bahadur could finally take five months of leave
and return home to begin reconstruction. He
said that rebuilding a small family house with
stone and mud mortar cost him around NPR
200,000. He and his mother said they had lost
hope that the government would assist them
and started rebuilding on their own. Their
house was categorized as partially damaged
despite being more damaged than some of their
neighbors’ houses which were categorized as
fully damaged in the CBS damage assessment.
Yagya Bahadur’s mother had filed a complaint
but feared the complaint would simply be treat-
ed as a formality and would not be addressed.
She thought that the assessment team had not
examined the houses properly: “My house is
listed in the partially damaged category, but
look, we are living in this shed!”
Soon, Yagya Bahadur would have to return
once more to Saudi Arabia to continue sending
money to his family in Prapcha and to pay back
the loans he took to pay for reconstruction.
62


Aid and Recovery in Post-Earthquake Nepal
Photo: Anurag Devkota
4.2 Sale of assets
Sales of assets was not common but those who
borrowed more frequently were also more
likely to have sold assets. Livestock remained
the most common asset sold, followed by land
and household goods.
Sales of assets remained low at 3% compared to 4%
in IRM-2. As before, people in severely hit districts
were more likely to sell assets than people in other
less affected districts. Among those who sold assets
in IRM-3, the majority of asset sales continued to be
of livestock (58%). Livestock sales were highest in the
severely hit districts with 87% of those who sold assets
in these districts saying they sold livestock. Land and
household goods were also commonly sold. In urban
areas, people were more likely to sell land (75% of
those who sold assets), while in rural areas sale of
livestock was more common (84%). Household goods
were more frequently sold in urban areas (16%) and
less remote regions (29%), compared to rural areas
(8%) or remote (4%) and more remote regions (5%).
People who borrowed more frequently were more
likely to have sold assets. According to results from
the IRM-2 and IRM-3 household panel dataset,
people who borrowed in both IRM-2 and IRM-3 were
7 percentage points more likely to sell assets to cope
with the earthquakes’ impacts.
63


Aid and Housing Reconstruction Cash Grants
The share of household identifying remit-
tances as a main income source grew over the
three research rounds but the number of those
reporting to have received them has declined.
Respondents in affected areas increasingly considered
remittances as a main source of income - Figure 4.8.62
Across all districts, 14% of people identified remit-
tances as a main income source in IRM-3 compared
to 9% in IRM-1. The share of households reporting
4.3 Migration and remittances
remittances as a main income source has increased
across each impact category since IRM-1.
However, the proportion of people receiving remit-
tances has declined slightly since IRM-2. Compared
to 21% in IRM-2, only 19% said they received remit-
tances in IRM-3. There have been large increases in
the proportion of people receiving remittances in
Nuwakot, Lamjung, and Okhaldhunga. Elsewhere, the
proportion receiving remittances has either declined
or stayed the same.
Figure 4.8: Remittances as a main income source - by district impact (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
The likelihood of receiving remittances did
not seem to be associated with level of housing
damage.
Where people were living in IRM-3 did not seem
to be associated with the likelihood of receiving
remittances. People living in their own houses, and
those who were living in shelters, were equally likely to
receive remittances. Similarly, there was not a strong
relationship between whether people were receiving
remittances and whether their income improved in
the three months prior to IRM-3 (September 2016).
62 Respondents could identify multiple main sources of income. As
such, percentages do not add up to 100%.
However, remittances were more likely to reach peo-
ple in rural and remote areas. However, there was a
slight decline in the proportion of people receiving
remittances in more remote regions in IRM-3. The
main beneficiaries of remittances were those with a
high pre-earthquake income. Twenty-four percent
of the high income group acknowledged receiving
remittances compared to 17% of those with medium
income and 17% in the low income group. Only 3% of
remittance flows had domestic origins.
The volume of remittances received remained
largely the same.
The vast majority of those who received remittances
before the earthquakes said the volume has stayed at
similar levels since the earthquakes (87%). A slightly
64


Aid and Recovery in Post-Earthquake Nepal
larger share of people in more remote regions said
that they had received more remittances since the
earthquakes. People in less remote and urban areas
were more likely to mention that they received less
since the earthquake.
Declines in remittance flows before and after the
earthquake were more likely to affect the poor and
the disabled (Table 4.9). Eleven percent of low income
individuals reported decreases while only 3% of high
income people who received remittances said the
same. Fifteen percent of people with a disability who
received remittances reported receiving less since the
earthquake compared to 7% people with no disability.
People living in temporary shelters or in neighbors’
houses were more likely to report a reduction in
remittances than those in their own house, although
those in neighbors’ houses were also more likely to say
remittances have increased.
Table 4.9: Changes in remittances since the earthquakes - by income and disability (IRM-3, weighted)
Receive less since the earthquake Receive similar level since the earthquake Receive more since the earthquake Refused/ Don’t know
Low income 11% 84% 4% 1%
Medium income 10% 84% 3% 3%
High income 3% 89% 3% 4%
No disability 7% 87% 3% 3%
Disability 15% 83% 2% 0%
Migration rates remained low but people in
severely hit districts were slightly more likely
to have migrated since the earthquakes. The
most common reasonfor migration were lack
of shelter, lack of livelihoods, and landslides.
Levels of migration since the earthquakes remained
low in IRM-3 at 3%. Since the earthquakes, people
in severely hit districts were slightly more likely to
have migrated. Eighty-five percent of these migration
cases took place before the 2015 monsoon, with 15%
occurring either during or after the monsoon. Nearly
86% of those who migrated later returned home.
The reasons stated for migrating in IRM-3 were dif-
ferent than those given in the IRM-2 survey. While
the most commonly cited reason for migration in
IRM-2 was a lack of shelter (66%), people in IRM-3
were more likely to report livelihoods problems as the
main reason (48%) - Figure 4.9. While most people’s
livelihoods were recovering (see Chapter 2), some have
not seen any improvements. For others, recovery had
started but was not well advanced leading people to
move to seek better opportunities. Almost one-quarter
of those migrating in IRM-3 reported landslides as a
reason.
Figure 4.9: Reasons for migration (IRM-2, IRM-3, weighted)
65


Aid and Housing Reconstruction Cash Grants
Many households said they were planning
to send at least one family member abroad
for work if they faced difficulties paying for
the reconstruction of their houses and to pay
back loans.
During the field research, many households pointed
out that they will likely have to resort to having family
members migrate for work to be able to repay loans
including the high amounts borrowed for housing
reconstruction (Case Study 4.2).
Case Study 4.2: Young entrepreneurs find new opportunities at home
Two young entrepreneurs from Ramechhap,
childhood friends Kul Bahadur Shrestha, 21
years old, and Tika Lal Shrestha, 22 years old,
were able to start new businesses in Ramechhap
municipality after the earthquakes. Kul Bahadur
used to work in Kathmandu but returned home
after the earthquakes. “I worked as a cook in
Boudha. My roommate died in the earthquake
and I was too scared to stay in Kathmandu so
I left the job and came back to the village,” he
explained. The two friends decided to invest
in poultry farming and later in a restaurant,
which they are operating in the market area of
Ramechhap municipality. Each of them had
to spend NPR 200,000 on the restaurant. But
they are satisfied as they earn NPR 50,000 per
month from the restaurant, excluding the NPR
3,000 rent. “The earthquake has affected a lot of
people but it has provided me a new chance to
explore my life. Now my life has been changed.
If there had been no earthquake, I would still be
working in Kathmandu,” said Kul Bahadur and
added, “I am very happy with this business.”
Unlike many other young men from Ramechhap,
Kul Bahadur and Tika Lal have no plans to go
abroad for work. “Now I don’t think of going
abroad, it is much better here,” said Tika Lal.
66


Aid and Recovery in Post-Earthquake Nepal
Photo: Nayan Pokharel
5.1 Roles of political parties in the provision of aid
With the decline in emergency relief and
the increasing focus on reconstruction, the
formal influence of political parties over the
coordination of assistance reduced.
In the early months after the earthquakes, political
parties played key roles in relief distribution through
the District Disaster Relief Committees (DDRCs) and
Relief Distribution Committees. But these bodies be-
came less influential as aid declined and the formal
influence of political parties over the coordination
of assistance also reduced significantly. The NRA’s
technical and bureaucratic approach to reconstruction
did not allow for a formal role for political parties in
the CBS damage assessment, cash grant agreement
and distribution processes, and various mechanisms
established to collect complaints about the assessment
and beneficiary lists. This was found to have signifi-
cantly reduced overall political party engagement in
earthquake-related activities compared to IRM-1 and
IRM-2.
The formation of District Coordination Committees
(DCCs) to coordinate and monitor reconstruction in
earthquake-affected districts, under the leadership of a
Member of Parliament from the same district, did not
lead to the more direct involvement of political party
leaders and members in the reconstruction process.63
The DCCs were found to be either dysfunctional or
ceremonial in the districts visited, with most of the
DCCs’ work being done by subcommittees led by dis-
trict government officials such as the Chief District
Officer or Local Development Officer. District-level
leaders across districts visited reported that they were
not invited to DCC meetings or the meetings of sub-
committees—in contrast to DDRC meetings in which
they are included regularly—nor to meetings of the
sub-regional NRA offices. DCCs in general had little
influence over reconstruction and the coordination
of aid and the cash grant process at the district level.
MPs, who are generally based in Kathmandu, have not
been able to regularly attend DCC meetings nor have
they remained informed on the specific needs and
challenges of reconstruction in their districts.
Local political leaders’ dissatisfaction rose
as they were increasingly sidelined from
the reconstruction process at the local level,
reducing their capacity to assist earthquake
victims.
Many local party representatives who were eager to
help locals in their recovery were frustrated by nation-
63 The Asia Foundation and Democracy Resource Center Nepal
(2016). Aid and Recovery in Post-Earthquake Nepal: Independ-
ent Impacts and Recovery Monitoring Phase 2 - Qualitative
Field Monitoring (February and March 2016). Kathmandu and
Bangkok: The Asia Foundation, p. 32.
67


Politics, Social Cohesion, and Conflict
Photo: Nayan Pokharel
al-level political disagreements and changes in recon-
struction policies, which delayed reconstruction and
caused uncertainty at the local level. This frustration
was reinforced as their exclusion from district-level
decision-making left them with little information on
the government’s reconstruction efforts and reduced
their ability to adequately provide information to their
communities and address the concerns of earthquake
victims. Local political parties considered the newly
formed local NRA offices and DCCs less suited to
understanding local contexts and political dynamics
than the DDRCs. They also felt that these new bodies
could not adequately represent the concerns of com-
munity members as they were led by ‘outsiders’ such
as government officials temporarily stationed in the
districts or by MPs based in Kathmandu.
Political activities returned to pre-earthquake
normalcy. Political parties’ role in local
governance remained the same.
Political parties returned to their usual activities
before the earthquakes, a trend that has continued
from IRM-2. Political parties were conducting few
programs specifically related to the earthquakes at the
local level, nor did they make comprehensive efforts
to form platforms to address issues related to recovery
and reconstruction. Individually, some political party
leaders were involved in the reconstruction process
but no evidence suggests that they were motivated
by political objectives. Political party interference in
post-earthquake aid was found to have decreased since
the end of the relief phase when direct aid distribution
declined significantly.
At the VDC level, dynamics between political par-
ties and government officials have remained largely
unchanged since the earthquakes with officials con-
tinuing to consult political party representatives for
decisions on local governance. There was no evidence
to suggest a significant change in political dynamics
at the local level even after political infighting and
the change of government at the center in mid-2016.
Leaders from different parties both cooperated and
policed each other through their involvement in infor-
mal All Party Mechanisms. In all districts, parties were
engaged in debates and discussions on the formation
of local bodies in their respective areas.64
Political party representatives were involved
in the cash grant agreement process. Initially,
they supported local obstructions of the cash
grant agreement process. However, they also
facilitated agreements between protesters
and government offices to resume the pro-
cess. In many areas, political parties then
68


Aid and Recovery in Post-Earthquake Nepal
informally assisted the cash grant agreement
process by helping individual earthquake vic-
tims andfacilitating communication between
government offices and local communities.
Obstructions and protests due to dissatisfaction with
the new CBS beneficiary lists often took the form of
community members and local political party repre-
sentatives and, in some cases, VDC officials visiting
district government offices coming together to demand
assurances that grievances would be addressed before
the grant agreement and disbursement moved ahead.
After assisting people to protest against the CBS
assessment and obstruct the cash grant agreement
process, political parties then actively negotiated
agreements with district officials on behalf of those
who protested against their exclusion from the new
beneficiary lists, and helped ensure that the cash grant
agreement process could resume.
After political parties informed communities that their
concerns would be addressed and the beneficiary lists
would be adjusted based on grievance forms submitted
to government offices, protests were called off and
the cash grant process was allowed to move ahead.
Political party representatives, especially at the VDC
level, then played important roles in facilitating the
cash grant process. In 10 out of the 12 VDCs visited
during the qualitative research where the cash grant
agreement process was underway, political parties
5.2
Dissatisfaction with political parties was
high. Communities generally said this was
because of their lack of involvement in recon-
struction rather than interference.
Like other aid providers, satisfaction with local polit-
ical parties has dropped - falling from 26% in IRM-2
to 21% in IRM-3. However, those who blamed political
parties for the unsatisfactory state of reconstruction
referred to their lack of engagement rather than po-
litical interference or bias as reasons. The informal
provided assistance. They informally helped VDC offi-
cials plan and coordinate the process, informed people
about the timing, procedures, and requirements for
signing cash grant agreements, and provided logistical
support such as helping victims fill in various forms,
submit grievance forms, and keeping the required
documents in order. Political parties’ informal facili-
tation was generally received positively by the people
and local government officials. Political parties were
often actively engaged in communication between
government offices and people in the villages on the
policies, rules, and procedures of the grant agreement
process and the disbursement of the grant as well as
building requirements. Political parties communicated
decisions of the VDC, such as the date and place for the
signing of reconstruction cash grant agreements, what
documents were required to conclude the agreement
process, and other relevant information.
Despite these findings from the qualitative field
research, data from the survey reveal that political
parties rank low among local sources providing in-
formation on aid. Only 7% said they received infor-
mation from them compared to WCFs (18%), VDC
secretaries (24%), and neighbors (82%). Furthermore,
despite more than half of people thinking their com-
munication with local political parties was good, their
satisfaction with the communication was not as high
as with security officers, local administrative centers,
and local community organizations.
Satisfaction with political parties
and future vote preferences
assistance provided during the cash grant agreement
process was welcomed by many but was not seen as
sufficient. People commonly blamed political parties
for the outcomes of damage assessments, procedural
hurdles to the conclusion of cash grant agreements,
what they perceived as inadequate assistance, and
even the unsatisfactory state of reconstruction. They
expected political parties to be more engaged in the
recovery and reconstruction process. As political par-
ties failed to meet these expectations, dissatisfaction
increased.
64 The 2015 Constitution of Nepal created three tiers of government:
local government, state government, and federal government.
In March 2016, the Government of Nepal constituted the
Commission for Restructuring of Village, Municipalities, and
Special, Protected and Autonomous Area, commonly known as
the Local Body Restructuring Commission (LBRC). The LBRC,
is currently in the process of determining the number and the
boundaries of local bodies in each district. In this process, the
LBRC initiated nationwide stakeholders’ consultations beginning
in late July 2016.
69


Politics, Social Cohesion, and Conflict
High dissatisfaction with political parties did
not lead to changes in which political party
people were supporting.
Communities mostly expressed dissatisfaction with
political parties in general rather than any one party
in particular. Dissatisfaction with political parties is
generally common in Nepal and people rarely held
specific local leaders accountable for the slow progress
of reconstruction.65
High levels of dissatisfaction do not seem to have led to
changes in which party people support. There was no
indication in any of the wards visited in the qualitative
fieldwork that people were changing or thinking of
changing who they would support. This finding cor-
responds with data from the survey on future voting
intention. There were no large changes in who people
said they will vote for in the next election. The vast
majority said they do not know (Figure 5.1).
While the majority of people who voted for any party in
the last elections were undecided as to who to vote for
next time around, those who choose a party preferred
the same party they voted for in the last elections
(Table 5.1). Just 1% of those who voted for the Nepali
Congress said they would vote for another party. The
figures were 3% for UCPN (Maoist-Centre), 1% for
RPP, and 0% for CPN-UML and RPP-N.
Figure 5.1: Voting preference for next election (IRM-1, IRM-2, IRM-3, weighted)
â–  IRM-1 â–  IRM-2 â–  IRM-3
Table 5.1: Future voting preferences - by past voting behavior (IRM-3, weighted)
If an election was to be held soon, which party would you vote for? |
Nepali Congress CPN-UML UCPN (Maoist- Centre) RPP-N RPP NMKP I will not vote Refused Don’t know
Nepali Congress 39% 1% 0% 0% 0% 0% 3% 2% 54%
CPN-UML 0% 36% 0% 0% 0% 0% 2% 3% 58%
UCPN (Maoist-Centre) 2% 1% 23% 0% 0% 0% 2% 2% 69%
Which political party RPP-N 0% 0% 0% 23% 0% 0% 0% 8% 69%
did you vote for in the RPP 1% 0% 0% 0% 38% 0% 0% 11% 49%
last elections? 1 did not vote 2% 2% 0% 0% 0% 0% 32% 2% 62%
NMKP 2% 0% 0% 0% 0% 31% 4% 13% 50%
Refused 0% 0% 0% 0% 0% 0% 1% 57% 42%
Don’t know 0% 0% 0% 0% 0% 0% 0% 1% 98%
Total 10% 7% 2% 0% 0% 0% 6% 10% 65%
65 The Carter Center, Local Governance in Nepal: Public Perceptions
and Participation, February 2014 (available at: https://www.
cartercenter.org/resources/pdfs/news/pr/nepal-o22814-local-
governance.pdf).
70


Aid and Recovery in Post-Earthquake Nepal
Photo: Binu Sharma
This may be at least in part due to the fact that no new
leadership has emerged that is challenging existing
political parties and no single political party has been
able to distinguish itself in the post-earthquake relief
and reconstruction processes. With political parties
enjoying de facto leadership status at the local level,
and in the absence of alternatives, people have no
option but to continue working with the traditional
political parties in their communities.
Nearly half of community members interviewed
during the qualitative research felt that local elec-
tions would improve the recovery and reconstruction
process. This reiterates the significance community
members attach to the role of political parties in their
VDCs and wards and the importance people attach
to holding local leaders accountable. Only a quarter
of community members thought that local elections
would be unlikely to affect reconstruction.66
66 In addition to key informant interviews, focus group discussions,
and informal conversations and observation, a minimum of 10
citizens were formally interviewed in each ward visited during
the qualitative research to directly collect the perspectives of
community members on the themes of the research. Citizen
interviews were done on the basis of a questionnaire with 20
questions covering the five themes of IRM research.
71


Politics, Social Cohesion, and Conflict
5.3 Security and crime
Most people felt safe. There was no change in
the proportion of people feeling safe between
the last two rounds.
Feelings of safety and security remained largely
unchanged in the last two rounds of the survey, but
have improved significantly from those observed in
IRM-1. Three percent of people in both IRM-2 and
IRM-3 reported feeling unsafe compared to 16% in
IRM-1. The proportion of people saying they feel very
safe increased between IRM-2 and IRM-3: from 54%
to 67% (Figure 5.2).
Figure 5.2: Perceptions of safety in the community - by district impact
(IRM-1, IRM-2, IRM-3, weighted)
Very safe
Somewhat safe
I Somewhat unsafe H Don't know
Very unsafe
There were declines in the proportion of people feeling
very safe in some districts. Amongst the severely hit
districts, Gorkha is the only district where the share
feeling very safe in their community declined since
the early weeks after the earthquake (74% IRM-1,
53% IRM-2, 60% IRM-3). Of the crisis hit districts,
people in Okhaldhunga were far less likely to feel very
safe now than in earlier surveys. Both of the hit with
heavy losses districts, Lamjung and Solukhumbu, saw
a decline in the share of people feeling very safe.
Remoteness, income, religion, and shelter
types affected perceptions of safety.
People in more remote regions in IRM-3 were more
likely to feel unsafe (7%) than those in remote (3%)
or less remote (2%) regions. Those in the low income
band (61%) were less likely to feel very safe than
those in the middle and high income bands (70%
each). Among religious groups, Christians (75%)
were the most likely and Buddhists (60%) the least
likely (Hindus 69%, Muslims 66%) to feel very safe
in their community. While most people said they feel
safe, those in self-constructed shelters on others’ land
were more likely to feel unsafe (Figure 5.3). This is a
continuation of the situation in IRM-2.67 There was
a small increase in the proportion of people feeling
unsafe who are in shelters on their own land.
67 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 2 - Quantitative Survey (February and March
2016). Kathmandu and Bangkok: The Asia Foundation, p. 120.
72


Aid and Recovery in Post-Earthquake Nepal
Figure 5.3: Perceptions of safety in the community -
by where people are living (IRM-2, IRM-3, weighted)
land
Very safe H Somewhat unsafe
Somewhat safe H Very unsafe
★Less than 1 % of the sample in this category
Table 5.2: Proportion of people reporting violence in their community - by district impact,
district and remoteness (IRM-1, IRM-2, IRM-3, weighted)
Violent incidents in community since the earthquake Violent incidents in community since the beginning of the monsoon Violent incidents in community since the end of the winter season
IRM-1 IRM-2 IRM-3
Severely hit 2.1% 0.4% 1.4%
Dhading 5.7% 0.3% 0.3%
Gorkha 0.9% 0.5% 1.2%
Nuwakot 2.2% 1.1% 2.6%
Ramechhap 0.0% 0.3% 1.0%
Sindhupalchowk 0.9% 0.0% 2.0%
Crisis hit 6.8% 0.6% 0.3%
Bhaktapur 4.5% 1.2% 0.0%
Kathmandu 7.5% 0.6% 0.3%
Okhaldhunga 2.0% 0.5% 0.7%
Hit with heavy losses 3.2% 0.4% 0.4%
Solukhumbu 1.7% 0.6% 0.0%
Lamjung 4.0% 0.3% 0.6%
Hit 2.0% 1.4% 0.9%
Syangja 2.0% 1.4% 0.9%
All districts 4.8% 0.6% 0.7%
Less remote 5.1% 0.4% 0.3%
Remote 0.3% 0.9% 1.0%
More remote 0.0% 0.7% 1.0%
73


Politics, Social Cohesion, and Conflict
As in earlier surveys, there were no notable differences
between feelings of safety among men and women,
those with a disability and those without, and among
different caste groups.
During the qualitative research, some people reported
feeling unsafe due to their housing condition. Many
of those interviewed for the qualitative research said
they felt unsafe because they still lived in damaged
houses with only minor repairs, because of the trauma
of living through the earthquakes, and because they
feared landslides. In Sindhupalchowk, in particular,
some women felt unsafe as they feared gender-based
violence, especially if they lived in temporary shelters.
Reported cases of violence were often linked to alcohol
abuse. It was difficult, however, to determine whether
alcohol consumption and gender-based violence have
increased since the earthquakes and how they are
changing over time.
As in past surveys, very few people reported
any violent incidents in their community.
Only 0.7% said there was a violent incident in their
community between IRM-2 and IRM-3 (0.6% in
IRM-2 and 4.8% in IRM-1). There was a notable,
albeit still small, increase in violence in Nuwakot and
Sindhupalchowk districts, both of which are severely
hit districts. However, violence in less remote areas
significantly declined in IRM-3 (by 5 percentage points
compared to IRM-1), while it increased slightly in
remote and more remote areas (by about 1 percentage
point compared to IRM-1) - Table 5.2.
Crime rates were more likely to have fallen than risen
after IRM-2. Less than 1% people felt that crime rose
between IRM-2 and IRM-3. Most (75%) said that
crime remained at the same level, while 21% said it
had fallen.
5.4 Trust and social cohesion
Social cohesion has generally been strong
since the earthquakes and social relations
remained largely unchanged between IRM-2
and IRM-3.
As was the case in the first year after the earthquakes,
social cohesion remained strong in most wards vis-
ited for the qualitative research. Local communities
worked together to construct temporary shelters in
the immediate aftermath of the earthquakes and con-
tinued to use shared labor practices to rebuild houses
or local infrastructure 18 months on. In IRM-3, there
were so no significant problems with social relations
in 31 out of the 36 wards studied. No new tensions
emerged in the 27 wards where social cohesion was
reported to have remained intact or had strengthened
in IRM-2. Among the nine wards where some tensions
or conflicts were reported in IRM-2, the situation had
not improved in five wards while in four wards local
tensions disappeared. A slight improvement in social
relations was therefore observed in IRM-3 compared
to IRM-2. Where social cohesion was poor, this was
attributed to tensions between local and resettled
communities and caste-based discrimination and
related conflicts.
Conflicts and tensions continued where lo-
cal disagreements over displacement and
resettlement had not been addressed. Water
shortages seemed to aggravate the situation.
Previous rounds of the qualitative research reported
cases of conflict or tensions around displacement
and resettlement in three locations: Prapcha VDC in
Okhaldhunga, Syaule VDC in Sindhupalchowk, and
Barpak VDC in Gorkha. These conflicts persisted as
there was no progress in finding long-term resettle-
ment solutions for the displaced.
Discrimination against Dalits was common and a
major factor leading to the emergence or continuation
of social tensions. Water scarcity and related conflicts
seemed to aggravate caste-based discrimination by
increasing tensions over access to water. Displaced
Dalits in Prapcha VDC in Okhaldhunga faced routine
acts of social discrimination and sometimes physical
acts of violence. In Prapcha, where drinking water was
scarce, caste-based discrimination led to a fight that
eventually had to be settled by the police. In Lisankhu
VDC in Sindhupalchowk, drinking water scarcity due
to the drying of water sources after the earthquakes
led to frequent verbal fights between women collecting
water (reported in IRM-2). By IRM-3, drinking
water facilities had been repaired and the number
of community water pipes had increased from 9 to
23, solving some of the previous tensions. However,
Dalits complained that they did not have enough
drinking water in their settlement and that they were
not allowed to collect water from the pipes in the other
groups’ settlement.
As the volume and coverage of aid declined,
complaints about and tensions related to un-
even access to aid and perceived discrimina-
tion in aid distribution also decreased.
74


Aid and Recovery in Post-Earthquake Nepal
Photo: Chiran Manandhar
Although findings from the survey demonstrate a
reduction in the number of people believing that aid
providers have distributed assistance fairly, the field
study found fewer complaints about other groups
unfairly receiving more aid and people were more
likely to say that they believed disadvantaged groups
should be given targeted aid. Cases of social tensions
around relief distribution that were reported in IRM-
2 had also disappeared by IRM-3, likely because the
volumes of aid had declined significantly.68
Strong social networks and social cohesion
facilitate recovery.
Extended social networks beyond the immediate
community or locality facilitated recovery after the
earthquakes. Practices of labor sharing were observed
in several wards since IRM-1. In IRM-3, labor sharing
to repair damaged homes or rebuild was observed
in three of the 36 wards visited for the qualitative
research and in several other locations, communities
raised money at their own initiative to repair local
infrastructure. Some also worked together to recover
common sources of livelihoods.
Access to credit, government offices, and aid was
often shaped by connections to wider social networks
beyond the immediate community. Some groups lack
access to such networks which has implications for
their recovery. Dalits were facing greater difficulties
accessing loans especially from formal sources but also
from moneylenders who tend to be high caste and to
discriminate against Dalits. While there were examples
of targeted aid, Dalits, other marginalized groups, and
those in very remote areas were generally finding
it harder to access information and resources from
formal sources of lending as well as government offices
and other distributors of aid. This has implications
for the vulnerability and long-term recovery of these
groups.
68IRM-2 reported that while social cohesion overall remained good,
ward residents often complained about unequal aid distribution.
Remarks included perceptions that those closer to the locations
where relief materials were dropped got more aid or that there
was some political appropriation or interventions diverting aid
to particular groups. The most common complaints were about
other ethnic and caste groups receiving more aid. However,
frustrations were more likely to be directed at decision-makers
and government offices than at other groups. Only in a few
cases did they turn into tensions within the wards. See, The Asia
Foundation and Democracy Resource Center Nepal (2016). Aid
and Recovery in Post-Earthquake Nepal: Independent Impacts
and Recovery Monitoring Nepal Phase 2 - Qualitative Field
Monitoring (February-March 2016). Kathmandu and Bangkok:
The Asia Foundation, p. 48.
75


Politics, Social Cohesion, and Conflict
5.5 Potential sources of conflict
Conflicts related to resettlement, water sourc-
es, and caste-based discrimination may esca-
late if these issues remain unresolved.
With many displaced people still living in temporary
shelters, and some clashes with local communities
observed since IRM-i, tensions between local and
displaced communities continue to be a source of
potential conflict, especially in the absence of clear
long-term resettlement plans. Displaced Dalits have
been the most likely to face discrimination and con-
flicts with local residents and this may be a cause for
escalating caste-based tensions. Delays in relocation
and geological assessments of the land of the displaced
increase the chance of such conflicts. Discontent and
conflicts often centered on the use of resources such
as water, land, and community forests. Research for
IRM-3 was conducted at the end of the monsoon sea-
son when water was plentiful. However, during the dry
winter season, water scarcity may intensify conflicts
within and between communities.
Frustrations of earthquake victims over the
slow pace of reconstruction andpolicy changes
may rise if assistance is delayedfurther. Such
dissatisfaction with the government and non-
governmental organizations may lead to new
conflicts or protests and violence.
Discontent over the slow pace of recovery and rebuild-
ing was high in most areas visited. While the cash
grant agreement process was being conducted and
some were beginning to receive the first instalment
of the cash grant at the time of the IRM-3 research,
many remained unsure whether and when they would
receive further assistance, especially in districts where
the cash grant agreement process had not yet begun in
September 2016. Many were also dissatisfied with the
assessments and process of identifying beneficiaries
and the number of official complaints was high in most
districts. Possible logistical delays in addressing these
complaints and informing victims of further steps in
the cash grant scheme may lead to tensions and pro-
tests in the districts.
While those excluded from beneficiary lists remain
uncertain about the government’s response to their
complaints, those who have already repaired or rebuilt
their houses were also unsure whether the houses
they have rebuilt would qualify retrospectively for the
reconstruction cash grant or future instalments of the
grant. As a political leader from Gorkha explained,
“In one VDC about 70 percent of people have rebuilt
their houses but they may not have necessarily used
earthquake-resistant techniques. The NRA provisions
state that rebuilt houses that are certified by engineers
as having followed the right techniques will also
qualify for the cash grant. But if the engineers do not
certify these houses, the people may beat and chase
them out of their villages.” The uncertainty and delay
in developing timely policy responses to such concerns
and communicating these clearly to the local level
point to the possibility of such issues giving rise to
conflict in the future.
76


Aid and Recovery in Post-Earthquake Nepal
6.1 Overview of conclusions
How have conditions evolved in the earthquake-af-
fected areas of Nepal? What are the key challenges
that need to be overcome if recovery is to take root?
And how can aid best support this? The Independent
Impacts and Recovery (IRM) project contributes in-
formation and analysis to help answer these questions
through longitudinal, mixed methods research.
This report outlines findings from the third round of
IRM research, conducted in September 2016. Combin-
ing findings from the survey and qualitative research,
it provides a snapshot of conditions almost eighteen
months on from the disasters. It makes comparisons
with data from past rounds of IRM to see how needs
and conditions are changing over time.
Progress in housing reconstruction remained
slow and many continued to stay in shelters.
Of those whose houses were seriously impacted by
the earthquakes, most had done nothing to repair or
rebuild. Some of those who returned to their houses
moved back into temporary shelters after realizing
that their damaged houses remained too unsafe or
were unsuited for living. Others stayed in their houses
despite the buildings remaining unsafe after only
minor repairs or being located on at-risk land.
As of September, 71% of people in the severely hit
districts were in temporary shelters. Alack of money,
and slow progress with the government’s flagship
housing reconstruction cash grant program, left many
people in shelters that they deemed to be inadequate.
The survey found that a large share of people struggled
to get their shelters ready for the monsoon, the second
they have faced since the disasters. Many got sick
during the monsoon due to issues with their shelters.
Slow progress was in large part due to delays and obsta-
cles in accessing the government’s housing reconstruc-
tion cash grants. These were common at the time of
research although much progress in distributing the first
installment of the grant have since been made. Howev-
er, the fact that few planned to use the first installment
of the cash grant for the intended purpose, and that
awareness of requirements, including building codes,
was low, indicates the potential for future problems.
Most cited a lack of money as the main reason
preventing themfrom starting to rebuild.
People said they urgently needed both money and
construction materials to be able to rebuild. Looking
forward, people were still concerned that the funds
they will receive under the housing reconstruction
cash grant program will cover but a small amount
of the costs needed for rebuilding. Costs for the
construction of houses were found to have increased
significantly due to higher prices for materials as well
as transportation and labor.
77


Conclusions and Recommendations
Photo: Anu rag Devkota
People received significantly less aid and
many of their pressing needs, especially those
beyond reconstruction, remained unidentified
and unaddressed.
The steep decline in the coverage of aid was true for
different types of assistance including relief, material
aid, and cash support. Only 15% of people have
received aid of any type since IRM-2. The drop in aid
coverage does not correspond with a decline in demand
for aid. IRM-3 found that people continued to have a
wide range of needs that were not being addressed
through government or non-government assistance.
On the contrary, as time passes, the gap between needs
and aid provided seemed to be increasing. As a result,
satisfaction with every aid provider reduced. Fewer
people thought that everyone can receive aid according
to their needs than in the past.
Borrowing has remained high and it will like-
ly increase further in thefuture.
The data show that marginalized populations—those
of low income, of low caste, the disabled, etc.—have
often borrowed repeatedly and at increasing volumes;
and this has not been associated with improvements in
their income, accommodation, or food consumption.
Informal sources of credit, from whom most people are
borrowing, often do not require collateral but charge
high interest rates. If debt-loads continue to increase,
some people may be stuck in situations where paying
off loans is impossible.
78


Aid and Recovery in Post-Earthquake Nepal
Peoplefrom marginalized and disadvantaged
groups were noticeably falling behind in their
recovery, becoming increasingly vulnerable.
This was particularly true for affected Dalits.
Lower caste and low income groups continue to face
the greatest challenges in recovering. They have had
higher rates of borrowing, mostly from informal
sources at high interest rates, leading to a risk of debt
traps. Inequality and prevalent forms of exclusion
and discrimination negatively affected the recovery
of Dalits who stood out as a highly vulnerable group
in this research round.
The displaced and those living in temporary shelters
remained among the most vulnerable groups facing
uncertainty and various risks. People from marginal-
ized groups were disproportionately likely to still be
living in temporary shelters and to be less prepared for
the monsoon. People in very remote areas were facing
greater obstacles to accessing aid and rebuilding their
houses. People with disabilities were also slower to re-
cover and women, children, and the elderly continued
to be seen as particularly vulnerable groups.
Landslides and water shortages continued to
be common and affect recovery.
Landslides were a common worry. Many people
were still waiting for geological land assessments to
determine whether or not their land was safe. Water
shortages, already reported in IRM-2, were still a
prominent concern due the drying of water sources,
damaged irrigation systems, and insufficient rainfall.
This particularly affected the recovery of farmers.
Communities were dissatisfied with commu-
nication about aid and resulting uncertainties.
Overall, people did not feel that they could communi-
cate well with aid providers, especially those removed
from the local level. The most common source for
information about aid were neighbors, radio, the VDC
office, and Ward Citizen Forums. Many highlighted
uncertainty resulting from a lack of clarity on time-
lines, procedures, and requirements of aid schemes,
including the housing reconstruction cash grant
scheme, as a pressing concern.
People continued to suffer psychological im-
pacts of the disaster.
Other issues, not explored in previous rounds of the
IRM survey, have also emerged. One-in-five people,
for example, report that someone in their household
has continuing trauma.
Despite the strain of recovery, some positive
trends were also observed: most notably,
social cohesion in communities remained
strong; and most were able to continue recov-
ering their livelihoods.
Social relations have remained strong after the earth-
quakes; crime has not increased; and violence contin-
ues to be rare. The quality of infrastructure and access
to public services has continued to improve. The need
for food declined in all districts although it continued
to be needed more in severely hit districts and remote
areas. The report also highlights that most people saw
their livelihoods recover further.
6.2 Key focus areas and recommendations
The data and analysis from the IRM-3 research has
established emerging challenges relevant to ongoing
and future assistance for earthquake recovery. The
National Reconstruction Authority (NRA), interna-
tional donors, and non-government organizations
have already begun to respond to some issues raised
by the research but challenges and risks remain. The
report concludes by providing a set of independent
recommendations for aid providers. The focus areas,
and the policy implications that flow from them, do
not necessarily reflect the views of the donors to the
project.69
Shelter and housing reconstruction
There is an urgent need to speed up the roll-out of
the cash grants through the housing reconstruction
program. Much progress in distributing the first
installment of the cash grant has been made since the
IRM-3 research was conducted but the survey data has
shown that needs remain great. That the cash grant
will likely cover but a small proportion of the costs for
families of rebuilding is worrying given that affordable
credit has not been made available in parallel. As a
result, people have to borrow large amounts from
69 These are independent recommendations rather than those of
the UK or Swiss government.
79


Conclusions and Recommendations
Photo: Binu Sharma
informal sources and at high interest rates, which has
already increased debt burdens. In addition, limited
awareness of and ability to fulfill requirements for
receiving subsequent installments of the grant may
mean that many will not receive further assistance.
The need for clearer information on, and assistance
with, procedures and building codes of the cash grant
scheme remain great, as does the need for additional
financial or material support to rebuild. The IRM-2
report also warned of the need to have a medium-term
strategy in place to improve the quality of temporary
shelters given that reconstruction will take time. The
authors believe this continues to be necessary.
The housing reconstruction cash grant scheme
Recommendation 4: Improve communication
between government offices by strengthening
coordination mechanisms and information flows
between the NRA and government line ministries
in Kathmandu, districts headquarters, and the
local level (rural municipalities or Gaupalika).
Roles and responsibilities of different bodies
need to be more clearly defined to improve
communication and coordination.
Recommendation 5: Develop plans for the
clear transfer of responsibilities related to
reconstruction and recovery work to new local
bodies after local body restructuring.
Recommendation 1: Information on proce-
dures of the government cash grant scheme needs
to be communicated quickly and more clearly to
local government offices and citizens. Local stake-
holders, who are close to affected communities,
should be utilized more for sharing information.
Recommendation 2: Information on challenges
related to accessing the grants after agreements
have been signed, as well as on the number of
people who have yet to withdraw the grant from
bank accounts, should be collected to improve
access for future rounds of grant dispersal.
Recommendation 3: Technical assistance
during reconstruction needs to be more widely
available across earthquake-affected districts.
Shelter conditions and displacement
Recommendation 6: Improve the quality
of existing shelters for the medium-term and
prioritize programs to mitigate the consequences
of staying in temporary shelter (targeted health
support and medicine, temporary water and
sanitation facilities, women’s security).
Recommendation 7: Complete assessments
to determine whether people can return to and
rebuild on land deemed to be at risk. Clearly
communicate the findings of such assessments to
local stakeholders and affected households.
Recommendation 8: Generate policy for sup-
porting the permanent resettlement of displaced
households unable to return to their land.
80


Aid and Recovery in Post-Earthquake Nepal
Debt and borrowing
Borrowing allowed people to overcome some of
the immediate challenges they have faced since the
earthquakes. But repeated borrowing of increasing
loan amounts is a cause for worry, especially given
that interest rates are climbing. While relatively few
have sold assets, either to raise funds or service exist-
ing debt, there is a risk of this in the future if people
cannot pay off debts or if their livelihoods do not fully
recover. Further cash grants, or the direct provision of
construction materials, are also needed to help people
overcome the earthquakes’ enduring impacts. Where
loans are provided, interest rates should be regulated.
This maybe particularly challenging given that access
to banks is much less common in the more-affected
remote areas and that disadvantaged groups face
specific challenges in accessing credit from formal
sources.
Recommendation 9: Expand soft loan pro-
grams, strengthen communication about them,
and ensure they reach those in remote areas and
marginalized groups.
Recommendation 10: Ensure better awareness of
government low interest loans in particular and
make these more widely available. Central-level
loan policies may need to be revised for ensure
better access for those in need of credit.
Needs beyond reconstruction
Many local needs remained unaddressed and there
was little shared understanding and coordination of
affected communities’ most urgent needs as well as
the specific needs of some groups. The report high-
lights that eighteen months after the disaster, a large
proportion of people continued to have psychosocial
problems that were triggered by the earthquakes or
by struggles since the disasters. Experiences from
other post-disaster contexts show that such problems
can last long after people economically get back on
their feet and back into their own houses. Tracking
trauma, and developing programs to respond to it,
is key. Further, many vulnerable and disadvantaged
groups had additional needs. Pre-existing needs of
poor households in rural Nepal have become more
pressing. For example, poor farmers were struggling in
their recovery as they were in greater need of financial
resources for rebuilding.
Recommendation 11: Strengthen communi-
cation channels for local communities to express
their needs.
Recommendation 12: Track long-term psycho-
social impacts of the earthquakes and their im-
plications for recovery and expand psychosocial
support for earthquake-affected communities.
Recommendation 13: Continue to provide
livelihood support to help generate incomes for
poor households, especially for farmers.
Making sure the marginalized do not get left behind
The IRM-3 data show strongly that some groups are
struggling more than others. The report finds sys-
tematic differences in the likelihood of moving back
to permanent housing, in livelihoods recovery, and in
decreases in food consumption between groups. Those
with a low income, no or little education, and those
with a disability are making the least progress. Low
income and low caste people are borrowing repeatedly
at increasing volumes but it appears that this is often
just to get by and is not leading to fuller recovery. Low
income people are far more likely to sell assets. The
evidence does not support the conclusion that the
struggles of these groups are a result of systematic
exclusion on the part of aid providers. Rather, these
groups face particular challenges, such as low capital
stocks and less well-remunerated job opportunities,
that make it harder for them to recover. Those strug-
gling tend to be the same people who were also most
vulnerable and marginalized before the earthquakes.
It is thus vital that more attention and resources are
directed to these groups so they are not left further
behind.
Recommendation 14: More attention needs
to be paid to the specific challenges of vulnera-
ble groups to facilitate special assistance that
enhances their ability to recover. This includes
the need to develop a greater understanding of
who is vulnerable in local areas and the factors
preventing vulnerable groups from rebuilding.
Recommendation 15: Targeted aid should be
context-sensitive; this means local communities
need to be informed of and involved in the
development and implementation of targeted aid
programs to avoid conflict.
81