Citation
Aid and Recovery in Post-Earthquake Nepal

Material Information

Title:
Aid and Recovery in Post-Earthquake Nepal Independent impacts and recovery monitoring phase 4, quantitative survey, April 2017
Creator:
The Asia Foundation ( Author, Primary )
UK aid ( contributor )
Swiss Development Cooperation ( contributor )
Interdisciplinary Analysts ( contributor )
Place of Publication:
San Francisco, CA
Publisher:
The Asia Foundation
Publication Date:
Copyright Date:
© 2017, 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 )
Politics ( SWAY )
Politics and government ( LCSH )
राजनीति ( SWAY )
National Reconstruction Authority ( SWAY )
Nepal. Rāṣṭriya punanirmāṇa prādhikaraṇa
राष्ट्रिय पुननिर्माण प्राधिकरण ( SWAY )
Elections (local) ( SWAY )
Nepal -- Local elections (local) ( LCSH )
स्थानीय चुनावहरु ( SWAY )
Genre:
NGO Report ( SWAY )
Temporal Coverage:
20150427 - 20170104
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:

00001.tiff

00002.tiff

00003.tiff

00004.tiff

00005.tiff

00006.tiff

00007.tiff

00008.tiff

00009.tiff

00010.tiff

00011.tiff

00012.tiff

00013.tiff

00014.tiff

00015.tiff

00016.tiff

00017.tiff

00018.tiff

00019.tiff

00020.tiff

00021.tiff

00022.tiff

00023.tiff

00024.tiff

00025.tiff

00026.tiff

00027.tiff

00028.tiff

00029.tiff

00030.tiff

00031.tiff

00032.tiff

00033.tiff

00034.tiff

00035.tiff

00036.tiff

00037.tiff

00038.tiff

00039.tiff

00040.tiff

00041.tiff

00042.tiff

00043.tiff

00044.tiff

00045.tiff

00046.tiff

00047.tiff

00048.tiff

00049.tiff

00050.tiff

00051.tiff

00052.tiff

00053.tiff

00054.tiff

00055.tiff

00056.tiff

00057.tiff

00058.tiff

00059.tiff

00060.tiff

00061.tiff

00062.tiff

00063.tiff

00064.tiff

00065.tiff

00066.tiff

00067.tiff

00068.tiff

00069.tiff

00070.tiff

00071.tiff

00072.tiff

00073.tiff

00074.tiff

00075.tiff

00076.tiff

00077.tiff

00078.tiff

00079.tiff

00080.tiff

00081.tiff

00082.tiff

00083.tiff

00084.tiff

00085.tiff

00086.tiff

00087.tiff

00088.tiff

00089.tiff

00090.tiff

00091.tiff

00092.tiff

00093.tiff

00094.tiff

00095.tiff

00096.tiff

00097.tiff

00098.tiff

00099.tiff

00100.tiff

00101.tiff

00102.tiff

00103.tiff

00104.tiff

00105.tiff

00106.tiff

00107.tiff

00108.tiff

00109.tiff

00110.tiff

00111.tiff

00112.tiff

00113.tiff

00114.tiff

00115.tiff

00116.tiff

00117.tiff

00118.tiff

00119.tiff

00120.tiff

00121.tiff

00122.tiff

00123.tiff

00124.tiff

00125.tiff

00126.tiff

00127.tiff

00128.tiff

00129.tiff

00130.tiff

00131.tiff

00132.tiff

00133.tiff

00134.tiff

00135.tiff

00136.tiff

00137.tiff

00138.tiff

00139.tiff

00140.tiff

00141.tiff

00142.tiff

00143.tiff

00144.tiff

00145.tiff

00146.tiff

00147.tiff

00148.tiff

00149.tiff

00150.tiff

00151.tiff

00152.tiff

00153.tiff

00154.tiff

00155.tiff

00156.tiff

00157.tiff

00158.tiff

00159.tiff

00160.tiff

00161.tiff

00162.tiff

00163.tiff

00164.tiff

00165.tiff

00166.tiff

00167.tiff

00168.tiff

00169.tiff

00170.tiff

00171.tiff

00172.tiff

00173.tiff

00174.tiff

00175.tiff

00176.tiff

00177.tiff

00178.tiff

00179.tiff

00180.tiff

00181.tiff

00182.tiff

SWAY001713_00001.pdf


Full Text
'â– >
*■•7


Aid and Recovery in
Post-Earthquake Nepal
Independent Impacts and Recovery Monitoring Phase 4
Quantitative Survey: April 2017


The Asia Foundation
UKaid
from the British people
Schwti CwifcaOralkm BviIHC
Ccmtedarawria Svixztns
Crtltaffarnzlllifi ayi/fa
Federal Department af Foreign Attain. FDFA
Swiss Agency far Development and Cooperation 5 DC
nflH ritcr-i* Mri’i




Aid and Recovery in
Post-Earthauake Neea
Independent Impacts and Recovery Monitoring Phase 4
Quantitative Survey: April 2017
-SflnwVohftclD
ConUKltmiztiiMJ
GonlAdortuiuni cvi/m
The Asia Foundation
UKaid
from the British people
Fedonal Deparlrneni nJ Foreign Affafra FDFA
Swiss Agwcy for Devefopmwt and Cooperation SDC
^4*4 It)tmfUH


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 4
Quantitative Survey: April 2017
© 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: Tanka Gurung
Design: Deddeaw Laosinchai


Aid and Recovery in Post-Earthquake Nepal
PREFACE
In 2015, two powerful earthquakes hit Nepal, killing
almost 9,000 people and displacing hundreds of
thousands more. Since then, The Asia Foundation
has been tracking how those affected by the earth-
quakes have recovered. Four rounds of research, con-
ducted at roughly six-month intervals, have provided
snap shots of conditions on the ground, including the
challenges people face, the aid they are receiving and
the extent to which they are coping.
This report presents findings from the fourth round of
research, which involved qualitative fieldwork and a
quantitative household survey in April 2017. Because
the same wards are visited in each round, with the
same people interviewed, the report gives an accurate
picture of how things have changed as time has passed.
The findings show there has been some progress in
supporting recovery. The incomes of most of those
affected by the earthquake have continued to recovery
and local markets are operating almost as normal.
Drops in food consumption, identified in earlier rounds
of research, are now less pronounced than before. The
disbursement of the first tranche of the government’s
housing grant has led some to start rebuilding.
Yet the reports also show the scale of the challenges
that remain. Two years on from the earthquakes, the
majority of those whose houses suffered major damage
or complete destruction remain in temporary shelters.
Rising construction costs have prevented many from
beginning to rebuild and people are increasingly bor-
rowing from informal lenders who charge high interest
rates. It is likely that many people will get stuck in a
debt trap, unable to repay the loans they have taken.
Most public infrastructure has not been rebuilt.
The reports also show a worrying divergence in the
experience of different groups; this requires urgent
policy attention. There are growing disparities in levels
of recovery among different socio-economic groups,
with many of the marginalized being left behind.
Those who had low incomes before the earthquakes,
e.g. Dalits, the disabled and widows, score lower
than others on most recovery indicators. Indeed, the
earthquakes appear to have exacerbated preexisting
inequalities. More needs to be done to help these
vulnerable groups.
We thank our research partners (Democracy Resource
Center Nepal and Interdisciplinary Analysts), our
donor partners (UK Department for International
Development and the Swiss Development Coopera-
tion), and Nepali government officials in the National
Reconstruction Authority and the Ministry of Federal
Affairs and Local Development for their support.
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
The IRM-4 survey was implemented by a team
from Interdisciplinary Analysts (IDA) led by
Sudhindra Sharma. While Sudhindra provided
overall guidance, Hiranya Baral coordinated the
survey fieldwork, Bal Krishna Khadka provided
essential support in thinking through the implications
of the technical aspects of the survey methodology,
Chandra KC worked on getting the dataset into a form
ready for analysis and generated a large set of initial
tables and Sandeep Thapa designed the software for
data entry. Kurt Burja of the World Food Programme
provided NeKSAP data.
Analysis of the data was done by Anup Phayal, Jui
Shrestha and Patrick Barron, who co-authored the
report. Sasiwan Chingchit provided research support
and inputs throughout.
A number of people provided useful inputs at various
stages, including in the formation of the question-
naires, and analysis of the data. They include George
Varughese and Lena Michaels (The Asia Foundation),
Sudip Pokharel and his team (Democracy Resource
Center Nepal) and the IDA team. Deddeaw Laosinchai
designs the report’s layout.
Many thanks to the people of the 11 affected districts
who spent time sharing their views with the research
teams. We particularly value the time they have taken
to contribute to the research.
The 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. Craig Irwin (UK DFID) and
Stefan Fuerst (SDC) have managed the project from
the donor side, and have given useful inputs at every
stage. The views here do not necessarily reflect the UK
or the Swiss government’s 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
F | lhis report provides findings from the fourth in a
I series of large-scale surveys, conducted in April
.X. 2017, two years on from the devastating earth-
quakes that hit Nepal. The report is part of the Inde-
pendent Impacts and Recovery Monitoring for Account-
ability in Post-Earthquake Nepal (IRM) project. Using
both quantitative surveying 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. The fourth survey involved
face-to-face interviews with 4,854 household respond-
ents in 11 districts. Stratified random sampling ensures
that those interviewed are representative of the wider
population in affected areas. Throughout the report,
fourth round survey data (IRM-4) are compared with
data collected in June 2015 (IRM-1), February-March
2016 (IRM-2) and September 2016 (IRM-3) to allow
for an assessment of changes over time.
Shelter
There has been limited progress since the earthquakes
in people moving from temporary shelters back into
their homes. Almost three-quarters of people in
earthquake-affected areas now live in their own homes
compared to 60% in the immediate aftermath of the
earthquakes. However, 62% of people in the severely
hit districts were still living in temporary shelters as
of April 2017. In Sindhupalchowk district, 84% of
people are still in shelters. Across all areas, almost
half of those whose house was completely destroyed
continue to live in temporary shelters. People in more
remote areas are far more likely to remain in shelters
than others. Those whose house was badly damaged or
destroyed in lesser affected districts have been much
more likely than those in severely hit districts to move
back home. Marginalized groups—those with a low
income, no education, the disabled, lower castes and
Janajatis and widows—are far more likely than others
to remain in temporary shelters.
There has been a decline in the number of people in
shelters that use tarpaulins or that are primarily built
from bamboo. A relatively higher share of people were
able to completely repair their shelters to be ready for
the winter in IRM-4 (14%) compared to IRM-3 (6%)
or IRM-2 (3%). But people in severely hit districts,
with a low income or from a low caste group, as well
as the disabled, were less likely to have their shelters
ready for the season.
Fifty-six percent of those whose house suffered
complete destruction or major damage reported that
they had done nothing to rebuild. Among those whose
house was completely destroyed or suffered major
damage, 62% in severely hit, 55% in crisis hit, 42% in
hit with heavy losses and 34% in hit districts have done
nothing to rebuild. People of low caste or low income
are less likely than others to have started rebuilding.
Those whose income has declined since the earthquake
are far less likely to have started rebuilding. Of those
who have started to rebuild, the largest share (21%)
began after the first monsoon before the first winter
after the earthquakes. Not having enough money is the
main reason (93%) for people not rebuilding, followed
by waiting for government grants.
Livelihoods, Food and Services
Over time, there has been a large drop in the num-
ber of people generating income through farming.
The proportion of people farming their own land
has dropped from 68% in IRM-1 to 53% in IRM-3
and IRM-4. Many more people are now generating
income through their own business or daily wage
work than in the past and remittances have become
more important. Most people continue to see im-
provements in their income sources but the propor-
tion seeing improvements in the past three months
has declined for most sources compared to IRM-3.
Daily wage work, business income and remittances
are the exceptions. By and large, incomes appear to
have recovered. Around one-third of people say their
current income is lower than before the earthquakes
but a significant proportion (27%) also say it has
increased.
V


Aid and Recovery in Post-Earthquake Nepal
People in severely hit and crisis hit districts and people
remaining in temporary shelter are more likely to have
seen a decline in income compared to those in lower
impact districts. Income recovery in more remote
areas is lagging behind that in other regions. People
who sustained greater damage to their house are also
more likely to struggle with income recovery. Those
who were poorer before the earthquake, or who come
from less privileged social groups, are much less likely
to have seen their income recover than others.
Only 7% of the population in IRM-4 say that food is
one of their most important immediate needs, down
from 27% in IRM-1. Stated need for food is higher for
those of low caste, those who had a low pre-earthquake
income and people who live in more remote areas.
Food prices appear to have increased more drastically
in more remote areas and in higher impact districts: an
average of 69% people in the top two impact categories
say that food prices have become much higher
compared to 47% in the lower two impact categories.
There do not appear to be widespread decreases in
food consumption.
Reported access to clean drinking water has declined
in severely hit and crisis hit districts, especially in
Gorkha and Nuwakot. Satisfaction rates with public
services have declined in IRM-4 with the exception
of electricity for which more people are satisfied than
in the past. Highest levels of dissatisfaction are with
drinking water (23%) and roads (15%). Low caste
people are more likely to be dissatisfied with drinking
water than others.
Coping Strategies
Borrowing continues to increase in affected districts.
Borrowing has risen most sharply in more affected
districts. Fifty-five percent of people have borrowed
in the last eight months in the severely hit districts,
compared to 24% in the early months after the earth-
quakes. A larger proportion of people in more remote
areas are borrowing than elsewhere. As in previous
surveys, those who had a low income before the earth-
quake and individuals of low caste are also more likely
to borrow than others. Borrowing in IRM-4 has also
increased among people with disabilities. People who
sustained greater damage to their house are also more
likely to borrow, and they are more likely than others
to borrow for rebuilding purposes. People in more
remote areas are borrowing from informal sources,
such as moneylenders, friends, relatives, neighbors
and other individuals, which typically charge higher
interest rates. In contrast, people in less remote areas
are borrowing more from formal sources. A higher
share of people in higher impact districts and more
remote areas are regular borrowers. They are also
more likely to say they will borrow in the near future.
Those in more remote regions, and in more affected
areas, are at greater risk of falling into debt traps.
While only 4% of people said they sold assets in IRM-2,
and 3% in IRM-3, 6% now report having sold assets in
the last eight months. Sales of assets remain highest
in the severely hit districts. People who sold assets
in IRM-4 were most likely to have sold land (43%
of those who sold assets) or livestock (40%). Data
confirm the earlier finding that borrowing frequency
is associated with the likelihood of asset sales. Those
who have borrowed repeatedly since February-March
2016 (IRM-2) are more than twice as likely as those
who have not borrowed in any of the last three waves
of the survey to sell assets. A slightly higher proportion
of people living in shelters sold assets.
Remittances are becoming more important as a source
of income. Fifteen percent of people in affected areas
say remittances are one of their main income sources
in IRM-4, compared to 9% in IRM-1. However,
remittances still tend to be more important in less
affected districts and for those with a high income.
Overall, 65% of people say migration levels have
remained the same, 20% say they have increased, and
4% say levels have decreased since the earthquakes.
Plans for migration in the next year suggest the
earthquakes have an influence as a majority of those
who plan to do so are from severely hit districts.
Earthquake Aid
The share receiving aid has gone up by 25 percentage
points compared to IRM-3 with 40% saying they have
received aid since September 2016. This is largely
due to the distribution of the first tranche of the
government’s housing reconstruction grant. Recent
aid distribution has been concentrated in the districts
that were severely or crisis hit and in remote and
more remote areas. The poor are more likely to have
received aid than others. Similar shares of men and
women, and those with and without a disability, have
received aid. The government has been the foremost
aid provider since the earthquake, and is almost the
sole provider of material aid since winter 2016. Cash
has been the most common form of assistance.
Those who received cash assistance from the govern-
ment have received on average NPR 56,845 to date;
those who received it from non-governmental sources
have got NPR 13,082. Cash is cited as the most needed
aid followed by reconstruction materials. Mention of
cash as a need has been growing steadily: 38% said
it was a current need in IRM-1 while 64% said it will
be needed in the near future in IRM-4. Despite more
aid going to more affected districts and more remote
areas, and to the poor, needs continue to be greater in
these places and for these people.
vi


Aid and Recovery in Post-Earthquake Nepal
Satisfaction with most aid providers plunged after
February 2016 and has stayed at similar levels since
then. People express the lowest levels of satisfaction
with local political parties, religious groups and private
businesses. Those in the severely hit districts have
been the most likely to think that aid distribution has
been fair in all four surveys and the share of people
believing so has remained stable. People with higher
incomes are less likely than those with lower incomes
to think that aid distribution has been fair. Most
people who think aid distribution has been unfair
believe that those belonging to lower castes are unable
to receive aid equally and according to their needs.
Lower caste people think they are more likely to be
treated unfairly by a wide margin: 64% compared to
39% of those of high castes and 36% of Janajatis.
More than 70% of people mentioned neighbors as
their prime source of aid information in both IRM-3
and IRM-4. People with higher incomes, and those
belonging to higher castes, are less likely than others to
say that neighbors are their top source of information
on aid. People think that communication with most
aid providers is either bad or okay; few say that
communication with aid providers is good.
National Reconstruction Authority
Assistance
People in severely hit districts are far more likely than
those in crisis hit districts to report that a Central Bu-
reau of Statistics assessment team came to their home.
According to respondents, nearly all houses in severely
hit districts have been classified as fully damaged, and
hence are eligible for the RHRP grant, with far fewer
houses classified this way in other districts. Most people
are satisfied with how their house was classified. Those
in hit with heavy losses and hit districts are more likely
to be dissatisfied as are those in less remote areas, lower
castes and those with a medium or high pre-earthquake
income. People whose house was classified as partially
damaged are the most likely to not be satisfied.
The first tranche of the Rural Housing and Reconstruc-
tion Program (RHRP) grant was received by nearly
everyone who said they were declared eligible for it.
The lowest coverage levels were in Kathmandu (81%)
and Dhading (86%). The severely hit districts have the
highest share of people who say they were declared
ineligible who feel they should have been eligible (82%
of those declared ineligible). Over seven in 10 people
declared ineligible in Okhaldhunga say they should
have been eligible. Thirty-three percent of those
declared ineligible, who say they should have been
eligible, say their house was only partially damaged.
Only around four in 10 said they would use the grant
to build a house following NRA guidelines. Many
say they will use it to pay off loans or for livelihoods.
Knowledge of grant requirements does not affect
intended use of the money. Recipients of the first
tranche of the grant generally found the process to
be easy. People are generally confident of getting
the second tranche, irrespective of how they have
used, or will use, the first tranche. Receiving the first
tranche does not necessarily translate into people
starting rebuilding. Fifty-eight percent of those who
received the first tranche have done something to start
rebuilding compared to 68% of people who have not
received the grant. Only 39% of people are aware of
the retrofitting program.
Illness and Trauma
More people fell sick in the winter than in the monsoon
that preceded it. Nineteen percent had a sickness in
their family in the winter. Sickness in IRM-4 was
most common in Dhading and Gorkha (27% each),
Okhaldhunga (24%) and Sindhupalchowk (22%).
Recurrent colds (33%), fevers (33%) and prolonged
colds (21%) were the most common illnesses. Those
with lower incomes were more likely to have had
someone in their family fall ill. People in temporary
shelters, particularly cowsheds, were more likely to
have fallen ill. Housing preparedness for adverse
weather greatly affected whether people fell ill.
Incidences of illness were highest among those unable
to make any repairs to their shelters and those who
made repairs that were not sufficient.
The number of people reporting that a family member
is suffering psychological effects from the earthquakes
has decreased. Fifteen percent of households now
report enduring psychological effects. Women (16%)
are slightly more likely than men (13%) to report psy-
chological effects. Those whose house was completely
destroyed, who are not living in their own house or
who had a low pre-earthquake income are more likely
to reporting enduring psychological effects.
Politics and Local Elections
Dissatisfaction with the role of political parties in
assisting recovery remains high. Fifty-nine percent of
people in all affected districts expressed dissatisfaction
with local political parties’ assistance with disaster re-
lief since September 2016. Forty-five percent of people
are dissatisfied with the role of local administrations
in disaster relief since the last monsoon. People in
Sindhupalchowk and Kathmandu, in more remote
regions, and those with higher socio-economic status
are the most likely to be dissatisfied with both political
parties and local administrations. Those who feel the
VDC/municipality distributed aid fairly are almost
twice as likely as others to be satisfied with political
vii


Aid and Recovery in Post-Earthquake Nepal
parties. With local elections approaching, reports of
visits of elected officials increased compared to Sep-
tember 2016 but remain lower than in the immediate
aftermath of the earthquakes.
When asked about the most important factors when
choosing who to vote for in the upcoming local elec-
tions, 67% favored candidates/parties that they per-
ceived would support local development, 30% men-
tioned that they would choose a candidate/party that
their family has always voted for and 25% mentioned
that they would support a candidate/party in line with
the choice of their friends. Reconstruction and recov-
ery of earthquake-affected areas was the next most
cited factor (20%). People in more remote regions, of
low caste or with a low income were more likely to pri-
oritize earthquake reconstruction and recovery when
making their voting choice. Four in 10 people in the
earthquake-affected areas said that they thought the
upcoming elections would be free and fair, but three
in 10 were unsure. Forty-two percent of people said
that reconstruction would work the same way as before
after the elections, but people were more likely to be
optimistic (4% much better, 30% much better) than
pessimistic (1% somewhat worse, 1% much worse).
Security and Social Relations
In the immediate aftermath of the earthquake, a rel-
atively high proportion of people said they felt either
very or somewhat unsafe. However, in subsequent
rounds of the survey, a negligible share of people have
said they felt unsafe with the exception of Kathmandu
and Syangja (8% and 7%, respectively). Similar shares
of men and women and of different caste groups have
felt unsafe. Few report a violent incident in their com-
munity. Kathmandu residents are the most likely to
have seen a violent incident in their community since
the earthquakes.
Most people say that you need to be careful in dealing
with other people; few say most people can be trusted.
Cooperation in times of an emergency, however, is
very likely. More people in the severely hit and crisis
hit districts now think it is very likely that people will
reduce their use of or share resources if an emergency
occurred in their community. Of groups different from
themselves, people trust those from a different area
the least; levels of trust in those from a different caste
or a different religion are similar. People belonging to
lower castes are less likely to think that cooperation
in their community is possible. Over seven in 10 say
that relations with neighbors have remained the same
since the earthquakes.
viii


Aid and Recovery in Post-Earthquake Nepal
LIST OF ACRONYMS
CA Constituent Assembly
CBS Central Bureau of Statistics
CGI Corrugated Galvanized Iron
EA Enumeration Area
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 IRM research (June 2015)
IRM-2 Second round of IRM research (February - March 2016)
IRM-3 Third round of IRM research (September 2016)
IRM-4 Fourth round of IRM research (April 2017)
MP Member of Parliament
NeKSAP Nepal Food Security Monitoring system
NGO Non-governmental organization
NPR Nepali Rupees
NRA National Reconstruction Authority
PDNA Post-Disaster Needs Assessment
RHRP Rural Housing Reconstruction Program
SLC School leaving certificate
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 XII
LIST OF TABLES XVI
Chapter 1. Introduction 1
1.1 Background 1
1.2 Methodology and approach 3
Chapter 2. Shelter 7
2.1 Where are people living? 9
2.2 Movements from shelter to houses 13
2.3 Quality of temporary shelters 15
2.4 Preparedness for adverse weather 17
2.5 Rebuilding and reconstruction 20
Chapter 3. Livelihoods, Food and Services 25
3.1 Recovery of livelihoods 26
3.2 Food 31
3.3 Public services 37
Chapter 4. Coping Strategies 41
4.1 Borrowing 43
4.2 Assets sales 58
4.3 Remittances 62
4.4 Migration 65
Chapter 5. Earthquake Aid 67
5.1 Aid coverage 68
5.2 People’s needs in earthquake-affected areas 75
5.3 Satisfaction with aid distribution 79
5.4 Aid communication 83
Chapter 6. National Reconstruction Authority Assistance 87
6.1 The Central Bureau of Statistics damage assessment 88
6.2 The Rural Housing Reconstruction Program (RHRP) grant 93
6.3 Retrofitting grant 105
6.4 Communication with the NRA 106
Chapter 7. Illness and Trauma 107
7.1 Illness due to problems with shelter 108
7.2 Psychological effects from the earthquakes 111
Chapter 8. Politics and Local Elections 115
8.1 Satisfaction with political parties and local administrations 117
8.2 Have elected officials visited earthquake-affected areas? 122
8.3 Factors determining voting choices 123
X


Aid and Recovery in Post-Earthquake Nepal
8.4 Will the local elections be free and fair? 126
8.5 Outlook on earthquake reconstruction work as a result of the local elections 128
Chapter 9. Security and Social Relations 131
9.1 Safety and security 132
9.2 Social cohesion 135
Chapter 10. Conclusions 141
Annex A Methodology 145
Sampling frame and district selection 145
Selection of VDC/wards within districts and replacement of VDC/wards 146
Selection of enumeration areas within VDC/wards 146
Selection of households within EAs 147
Selection of respondents within households 147
Weighting data 147
Annex B Sample Characteristics 149
Gender 150
Caste 150
Income bands 151
Disability 151
Rural/urban areas and remoteness 152
Level of housing damage 152
Current type of shelter 153
xi


Aid and Recovery in Post-Earthquake Nepal
LIST OF FIGURES
Figure 2.1: Where people were/are living (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 9
Figure 2.2: Where people are living - by remoteness (IRM-4, weighted) 10
Figure 2.3: Share of people whose house was completely destroyed or suffered major 11
damage - by district impact (IRM-4, weighted)
Figure 2.4: Share of people living in temporary shelters (on own, public or other’s land) - 11
by level of house damage (IRM-4, weighted)
Figure 2.3: Share of people who were living in shelter in IRM-2 who moved to their 13
own house in IRM-4 (IRM-2, IRM-4, household panel, unweighted)
Figure 2.6: Share of people living in different types of shelters (IRM-2, IRM-3, IRM-4, weighted) 15
Figure 2.7: Share of people preparing their shelters for winter (IRM-2, IRM-4)/ 17
monsoon (IRM-3) (IRM-2, IRM-3, IRM-4, weighted)
Figure 2.8: Share of people preparing their shelters for winter - by caste (IRM-4, weighted) 18
Figure 2,9: Share of people preparing their shelters for winter - by pre-earthquake 19
income (IRM-4, weighted)
Figure 2.10: Share of people preparing their shelters for winter - by disability (IRM-4, weighted) 19
Figure 2.11: Time when respondents started to rebuild their house or build a new house 21
of those who have started to rebuild (IRM-4, weighted)
Figure 2.12: Proportion who have not done anything to rebuild their damaged house - 21
by caste, pre-earthquake income, education, gender, widows and disability
(IRM-4, weighted)
Figure 2.13: Change in income since the earthquake and delay in rebuilding house 22
(IRM-4, weighted)
Figure 2.14: Reasons for stopping repairing or not building a house (IRM-4, weighted) 22
Figure 3.1: Income sources for people in affected areas (IRM-1, IRM-2, IRM-3, IRM-4, 27
weighted)
Figure 3.2: Share of people within each income source whose income from that source 27
has improved (IRM-2, IRM-3, IRM-4, weighted)
Figure 3.3: Current income (IRM-4) compared to pre-earthquake income (IRM-2) 28
(IRM-4, IRM-2 household panel, unweighted)
Figure 3.4: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 29
by remoteness (IRM-4, IRM-2 household panel, unweighted)
Figure 3.5: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 29
by housing damage (IRM-4, IRM-2 household panel, unweighted)
Figure 3.6: Food as a top immediate need and three month need (IRM-1, IRM-2, 31
IRM-3, IRM-4, weighted)
Figure 3.7: Food as a top immediate need and three month need - by district impact 31
(IRM-4, weighted)
Figure 3.8: Food as a top immediate need and three month need - by remoteness 32
(IRM-4, weighted)
Figure 3.9: Increase in food prices - by district impact (IRM-4, weighted) 33
Figure 3.10: Increase in food prices - by remoteness (IRM-4, weighted) 34
Figure 3.11: Food consumption compared to last year (IRM-2, IRM-4, weighted) 34
Figure 3.12: Changes in food consumption in the past eight months (IRM-2, IRM-3, 35
IRM4, weighted)
Figure 3.13: Share saying they have services provided by the VDC/municipality 37
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Figure 3.14: Dissatisfaction with public services - by remoteness (weighted, IRM-4) 39
xii


Aid and Recovery in Post-Earthquake Nepal
Figure 4.1: Share of people who have borrowed - by remoteness (IRM-1, IRM-2, 44
IRM-3, IRM-4, weighted)
Figure 4.2: Share of people who have borrowed - by pre-earthquake income (IRM-2, 44
IRM-3, IRM-4, weighted)
Figure 4.3: Share of people who have borrowed - by caste (IRM-2, IRM-3, IRM-4, weighted) 45
Figure 4.4: Share of people who have borrowed - by disability (IRM-2, IRM-3, IRM-4, weighted) 45
Figure 4.5: Share of people who have borrowed - by housing damage (IRM-2, IRM-3, 46
IRM-4, weighted)
Figure 4.6: Share of people who have borrowed - by where people live (IRM-2, IRM-3, 46
IRM-4, weighted)
Figure 4.7: Reasons for borrowing, share of those borrowing (IRM-2, IRM-3, IRM-4, weighted) 47
Figure 4.8: Share of those taking loans who are borrowing for rebuilding - by housing 47
damage (IRM-4, weighted)
Figure 4.9: Average borrowing in NPR - by remoteness (IRM-1, IRM-2, IRM-3, 48
IRM-4, weighted)
Figure 4.10: Sources of borrowing among those who borrowed - by remoteness 50
(IRM-4, weighted)
Figure 4.11: Changes in interest rates from different sources (IRM-1, IRM-2, IRM-3, 51
IRM-4, weighted)
Figure 4.12: Changes in debt (IRM-4, weighted) 52
Figure 4.13: Increase in debt - by housing damage (IRM-4, weighted) 53
Figure 4.14: Unsuccessful borrowers - by district impact and remoteness (IRM-2, 55
IRM-3, IRM-4, weighted)
Figure 4.15: Unsuccessful borrowers - by caste and pre-earthquake income (IRM-2, 56
IRM-3, IRM-4, weighted)
Figure 4.16: Share of people who plan to borrow in the next three months (IRM-2, 56
IRM-3, IRM-4, weighted)
Figure 4.17: Share of people who intend to borrow in the next three months - by 57
remoteness (IRM-4, weighted)
Figure 4.18: Share of people who intend to borrow in the next three months - by gender, 57
widows, disability and housing damage (IRM-4, weighted)
Figure 4.19: Share of people who intend to borrow in the next 3 months - by caste and 58
pre-earthquake income (IRM-4, weighted)
Figure 4.20: Share of people who sold assets to cope with the earthquake impacts - 58
by district impact (IRM-2, IRM-3, IRM-4, weighted)
Figure 4.21: Share of people who sold assets to cope with the earthquake impacts - 59
by remoteness (IRM-2, IRM-3, IRM-4, weighted)
Figure 4.22: Number of time periods borrowed and selling of assets (IRM-2, IRM-3, 60
IRM-4 household panel, unweighted)
Figure 4.23: Share of people selling assets - by where people are living (IRM-3, 60
IRM-4, weighted)
Figure 4.24: Share of people selling assets - by pre-earthquake income (IRM-3, IRM-4, 61
weighted)
Figure 4.25: Remittances as a main income source - by remoteness (IRM-2, IRM-3, 62
IRM-4, weighted)
Figure 4.26: Remittances as a main income source - by pre-earthquake income 63
(IRM-2, IRM-3, IRM-4, weighted)
Figure 4.27: Remittances as a main income source - by housing damage (IRM-2, 63
IRM-3, IRM-4, weighted)
Figure 4.28: Change in number of people migrating from respondents’ communities 66
compared to before the earthquake - by remoteness (IRM-4, weighted)
Figure 5.1: Share of people receiving some type of aid - by district impact 69
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Figure 5.2: Share of people receiving some type of aid since the end of winter 2016 - 69
by district (IRM-4, weighted)
Figure 5.3: Proportion who received aid - by remoteness (IRM-1, IRM-2, IRM-3, 70
IRM-4, weighted)
xiii


Aid and Recovery in Post-Earthquake Nepal
Figure 5.4: Proportion who received aid - by caste (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 71
Figure 5.5: Proportion who received aid - by pre-earthquake income (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 71
Figure 5.6: Source of aid amongst those who received aid (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 73
Figure 5.7: Share receiving cash from the government and non-governmental sources (IRM-2, IRM-3, IRM-4, weighted) 73
Figure 5.8: Change in the share of people who agree that VDCs/municipalities have been distributing aid fairly - by district impact (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted) 79
Figure 5.9: Opinions on whether all can get aid equally according to their needs (IRM-2, IRM-3, IRM-4, weighted) 81
Figure 5.10: Groups that people think tend to get less aid among those who say not everyone is able to get aid equally (IRM-2, IRM-3, IRM-4 weighted) 82
Figure 5.11: Groups that tend to get less aid among those who think not everyone is able to get aid equally - by caste (IRM-4, weighted) 82
Figure 5.12: Source of information on aid - by where people are living (IRM-4, weighted) 84
Figure 5.13: Ease of communication with aid providers (IRM-3, IRM-4, weighted) 86
Figure 6.1: Whether the official damage assessment team visited - by remoteness (IRM-4, weighted) 89
Figure 6.2: Satisfaction with official housing damage classification - by district impact (IRM-2, IRM-3, IRM-4 household panel, unweighted) 91
Figure 6.3: Dissatisfaction with housing damage assessment - by remoteness (IRM-4, weighted) 92
Figure 6.4: Dissatisfaction with housing damage assessment - by pre-earthquake income (IRM-4, weighted) 92
Figure 6.5: Dissatisfaction with housing damage assessment - by housing damage classification (IRM-4, weighted) 92
Figure 6.6: Ineligible people who think they should have been eligible - by caste, gender, widows and pre-earthquake income (IRM-4, weighted) 94
Figure 6.7: Ineligible people who think they should have been eligible - by self-reported level of housing damage (IRM-4, weighted) 94
Figure 6.8: Awareness of increase in RHRP grant to NPR 300,000 - by grant eligibility and having received the grant (IRM-4, weighted) 97
Figure 6.9: Awareness of increase in RHRP grant to NPR 300,000 - by caste, gender and widows (IRM-4, weighted) 97
Figure 6.10: Knowledge of requirements to get the second tranche of RHRP grant among those who received the first tranche - by gender, widows and caste (IRM-4, weighted) 98
Figure 6.11: Difficulty of receiving the first tranche of the NRA housing grant - by gender, caste, widows (IRM-4, weighted) 99
Figure 6.12: Difficulty of receiving the first tranche of the NRA housing grant - by pre-earthquake income (IRM-4, weighted) 99
Figure 6.13: Confidence in getting the second tranche of the RHRP grant among those who got the first tranche - by remoteness (IRM-4, weighted) 100
Figure 6.14: Confidence in getting the second tranche of the NRA grant among those who got the first tranche - by gender, caste and widows (IRM-4, weighted) 101
Figure 6.15: Confidence in getting the second tranche of the RHRP grant among those who got the first tranche - by ease of receiving first tranche (IRM-4, weighted) 101
Figure 6.16: Confidence in getting the second tranche of the RHRP grant among those who got the first tranche - by ratings of communication with the NRA (IRM-4, weighted) 102
Figure 6.17: Awareness of retrofitting grant - by caste, gender and widows (IRM-4, weighted) 105
Figure 6.18: Communication with NRA - by whether or not RHRP grant was received (IRM-4, weighted) 106
xiv


Aid and Recovery in Post-Earthquake Nepal
Figure 7.1: Share of people who say someone in their family got sick due to shelter 109
issues - by where people were living (IRM-3, IRM-4 household panel, unweighted)
Figure 7,2: Share of people who say someone in their family got sick due to shelter 109
issues - by type of shelter (IRM-3, IRM-4 household panel, unweighted)
Figure 7.3: Share of people who say someone in their family got sick due to shelter 110
issues - by how ready their shelter was for adverse weather (IRM-3, IRM-4
household panel, unweighted)
Figure 8.1: Satisfaction with local political parties since the last monsoon - 117
by remoteness (IRM-4, weighted)
Figure 8.2: Satisfaction with local administrations since the last monsoon - 118
by remoteness (IRM-4, weighted)
Figure 8.3: Share saying elected officials have not visited their area in the previous six 122
months (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted)
Figure 8.4: Share unsure who to vote for - by whether elected officials visited area 123
(IRM-4, weighted)
Figure 8.5: Outlook on earthquake reconstruction work as a result of the local elections - 130
by whether people received RHRP grant for housing (IRM-4, weighted)
Figure 9.1: Perceptions of security - by district impact (IRM-1, IRM-2, IRM-3, IRM-4 132
household panel, unweighted)
Figure 9.2: Share feeling either very or somewhat unsafe and insecure in their 133
community - by gender (IRM-1, IRM-2, IRM-3, IRM-4 household panel,
unweighted)
Figure 9.3: Share feeling either very or somewhat unsafe and insecure in their community - 134
by caste (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted)
Figure 9.4: Share trusting different groups of people (IRM-3, IRM-4 household panel, 135
unweighted)
Figure 9.5: Share trusting people of a different caste - by caste (IRM-3, IRM-4 136
household panel, unweighted)
Figure 9.6: Share trusting people of a different religion - by religion (IRM-3, IRM-4 136
household panel, unweighted)
Figure 9.7: Likelihood of people in the community conserving food or water if 137
asked by the government in case of an emergency - by district impact
(IRM-2, IRM-3, IRM-4 household panel, unweighted)
Figure 9.8: Share saying people in their community would be very likely to cooperate - 138
by caste (IRM-2, IRM-3, IRM-4 household panel, unweighted)
Figure 9.9: Relations with neighbors after the earthquake - by district impact 139
(IRM-3, IRM-4 household panel, unweighted)
Figure A.i: Distribution of sample 146
Figure A.2: Distribution of sampled wards in 11 districts 147
xv


Aid and Recovery in Post-Earthquake Nepal
LIST OF TABLES
Table 1.1: Districts surveyed (IRM-4) 3
Table 2.1: Where people are living now - by district impact and district 9
(IRM-4, weighted)
Table 2.2: Share of people with a completely destroyed or majorly damaged house 11
living in shelters - by district impact and district (IRM-4, weighted)
Table 2.3: Share of people living in temporary shelters (on own, public or other’s land) - 12
by caste, pre-earthquake income and education (IRM-4, weighted)
Table 2.4: Share of people living in temporary shelters (on own, public or other’s land) - 12
by gender, widows and disability (IRM-4, weighted)
Table 2.5: Share of people living in shelters in IRM-2 who continue to live in shelters 14
in IRM-4 - by district impact and district (IRM-2, IRM-4, household
panel, unweighted)
Table 2.6: Share of people living in shelters in IRM-2 who continue to live in shelters 14
in IRM-4 - by caste, pre-earthquake income, education, gender, widows
and disability (IRM-2, IRM-3, IRM-4, household panel, unweighted)
Table 2.7: Share of people living in different types of shelters - by district impact, 15
district and remoteness (IRM-4, weighted)
Table 2.8: Share of people living in different types of shelters - by caste, pre-earthquake 16
income, education, gender, widows and disability (IRM-4, weighted)
Table 2.9: Share of people preparing their shelters for winter - by district impact, 18
district and remoteness (IRM-4, weighted)
Table 2.10: Proportion whose house was destroyed or suffered major damage who 20
have done nothing to rebuild their damaged house - by district impact
and district (IRM-4, weighted)
Table 2.11: Proportion who have stopped the rebuilding process or not built a house 23
because waiting for government cash grants - by district impact and
remoteness (IRM-3, IRM-4, weighted)
Table 2.12: Reasons for stopping the rebuilding process or not building a house - 23
by caste, pre-earthquake income, gender, widows, disability and housing
damage (IRM-4, weighted)
Table 3.1: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 28
by district impact and district (IRM-4, IRM-2 household panel, unweighted)
Table 3.2: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 29
by where people are living (IRM-4, IRM-2 household panel, unweighted)
Table 3.3: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 30
by caste, pre-earthquake income, widows and disability (IRM-4, IRM-2
household panel, unweighted)
Table 3.4: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - 30
by education (IRM-4, IRM-2 household panel, unweighted)
Table 3.5: Food as a top immediate need and three month need - by district impact 32
and district (IRM-4, weighted)
Table 3.6: Food as a top immediate need and three month need - by caste and 32
pre-earthquake income (IRM-4, weighted)
Table 3.7: Food as a top immediate need and three month need - by widows and 33
disability (IRM-4, weighted)
Table 3.8: Changes in food consumption in the past eight months - by district impact, 35
district and remoteness (IRM4, weighted)
XV i


Aid and Recovery in Post-Earthquake Nepal
Table 3.9: Changes in food consumption in the past eight months - by caste, 36
pre-earthquake income and gender (IRM4, weighted)
Table 3.10: Changes in food consumption in the past eight months in Okhaldhunga, 36
Sindhupalchowk, Nuwakot and Dhading - by caste and pre-earthquake
income (IRM4, weighted)
Table 3.11: Access to clean drinking water - by district impact and district (IRM-1, 37
IRM-2, IRM-3, IRM-4, weighted)
Table 3.12: Satisfaction with public services (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 38
Table 3.13: Dissatisfaction with public services - by district impact and district 39
(IRM-4, weighted)
Table 3.14: Dissatisfaction with public services - by caste and pre-earthquake income 40
(IRM-4, weighted)
Table 3.15: Dissatisfaction with public services - by gender, widows and disability 40
(Apr 2017, IRM-4, weighted)
Table 4.1: Share of people who have borrowed - by district impact and district 43
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Table 4.2: Share of people who have borrowed - by occupation (IRM-2, IRM-3, 45
IRM-4, weighted)
Table 4.3: Reasons for borrowing, share of those borrowing - by district impact 47
(IRM-4, weighted)
Table 4.4: Average borrowing in NPR - by district impact and district (IRM-1, IRM-2, 48
IRM-3, IRM-4, weighted)
Table 4.5: Sources of borrowing among those who borrowed (IRM-1, IRM-2, IRM-3, 49
IRM-4, weighted)
Table 4.6: Average borrowing in NPR - by sources (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 50
Table 4.7: Mean reported interest rates - by district impact, district and remoteness 51
(IRM-4, weighted)
Table 4.8: Changes in debt - by district impact, district and remoteness (IRM-4, weighted) 52
Table 4.9: Overall debt - by caste and pre-earthquake income (IRM-4, weighted) 53
Table 4.10: Overall debt - by borrowing frequency (IRM-2, IRM-3, IRM-4 panel, unweighted) 53
Table 4.11: Borrowing frequency - by housing damage (IRM-2, IRM-3, IRM-4 panel, 54
unweighted)
Table 4.12: Borrowing frequency - by caste and pre-earthquake income (IRM-2, 54
IRM-3, IRM-4 panel, unweighted)
Table 4.13: Borrowing frequency - by district impact, district and remoteness (IRM-2, 55
IRM-3, IRM-4 panel, unweighted)
Table 4.14: Share of people who plan to borrow in the next three months - by district 57
impact and district (IRM-4, weighted)
Table 4.15: Types of assets sold to cope with earthquake impacts amongst those who 59
sold assets - by district impact, district and remoteness (IRM-4, weighted)
Table 4.16: Remittances as a main income source - by district impact and district 62
(IRM-2, IRM-3, IRM-4, weighted)
Table 4.17: Share of people receiving remittances - by pre-earthquake income 64
(IRM-4, weighted)
Table 4.18: Share of people receiving remittances - by housing damage 64
(IRM-4, weighted)
Table 4.19: Share of people receiving remittances - by where people are living 64
(IRM-4, weighted)
Table 4.20: Change in number of people migrating from respondents’ communities
compared to before the earthquake - by district impact and district 65
(IRM-4, weighted)
Table 5.1: Change in aid coverage - by district and district impact (IRM-1, IRM-2, 70
IRM-3, IRM-4, weighted)
Table 5.2: Share saying they do not need aid either now or in the future - by district 72
impact and district (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
xvii


Aid and Recovery in Post-Earthquake Nepal
Table 5.3: Average cash amount received to date (NPR) from the government 74
Table 5.4: and non-governmental sources - by district impact and district (IRM-2, IRM-3, IRM-4, weighted) Most mentioned current needs - by district impact and district 75
Table 5.5: (IRM-4, weighted) Most mentioned current needs - by pre-earthquake income (IRM-4, weighted) 76
Table 5.6: Most mentioned current needs - by gender, widows and caste (IRM-4, weighted) 76
Table 5.7: Most mentioned current needs - by where people are living (IRM-4, weighted) 76
Table 5.8: Top anticipated needs in three months - by district impact and district 77
Table 5.9: (IRM-4, weighted) Top anticipated needs in three months - by pre-earthquake income (IRM- 77
Table 5.10: 4, weighted) Changes in current and anticipated needs (IRM-1, IRM-2, IRM-3, IRM-4, 78
Table 5.11: weighted) Proportion satisfied with aid providers (IRM-1, IRM-2, IRM-3, IRM-4, weighted) 79
Table 5.12: Change in the share of people who agree that VDCs/municipalities have 80
Table 5.13: been distributing aid fairly - by district impact and district (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted) Perceptions that aid distribution has been fair - by gender, caste and pre-earthquake 80
Table 5.14: income (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted) Top five sources of information on aid - by district impact and district 83
Table 5.15: (IRM-3, IRM-4, weighted) Top five sources of information on aid - by pre-earthquake income, age, 84
Table 5.16: disability, gender, women’s marital status and caste (IRM-4, weighted) Top five sources of information on aid - by pre-earthquake income, age, 85
Table 5.17: disability, gender, widows and caste (IRM-4, weighted) Satisfaction with how aid providers have communicated about aid 85
(IRM-3, IRM-4, weighted)
Table 6.1: Whether the official damage assessment team visited - by district impact and district (IRM-4, weighted) 89
Table 6.2: Results of the official damage assessment - by district impact and district (IRM-4, weighted) 90
Table 6.3: Satisfaction with the most recent housing damage assessment - by district and district impact (IRM-4, weighted) 90
Table 6.4: Share who received the first tranche of the RHRP grant among those who say they were declared eligible - by district impact and district (IRM-4, weighted) 93
Table 6.5: Ineligible people who think they should have been eligible - by district impact and district (IRM-4, weighted) 93
Table 6.6: Use of/planned use of first tranche of RHRP grant among those declared eligible to receive it - by district impact and district (IRM-3, IRM-4, weighted) 95
Table 6.7: Use of/planned use of first tranche of RHRP grant among those declared eligible to receive it - by pre-earthquake income (IRM-4, weighted) 96
Table 6.8: Use of first tranche of NRA grant - by knowledge of conditions for second tranche (IRM-4, weighted) 96
Table 6.9: Awareness of increase in NRA housing grant to NPR 300,000 - by district impact and impact (IRM-4, weighted) 97
Table 6.10: Knowledge of requirements to get the second tranche of RHRP grant among those who got the first tranche - by district impact and district (IRM-4, weighted) 98
Table 6.11: Ease of getting first tranche of NRA grant - by district impact and district (IRM-4, weighted) 99
Table 6.12: Confidence in getting the second tranche of the RHRP grant among those who got the first tranche - by district and district impact (IRM-4, weighted) 100
Table 6.13: Confidence in receiving second tranche of RHRP grant - by intended use/usage of first tranche (IRM-4, weighted) 102
Table 6.14: Whether people have started rebuilding homes - by whether or not they received the first tranche (IRM-4, weighted) 103
xviii


Aid and Recovery in Post-Earthquake Nepal
Table 6.15: Whether people who have received the first tranche of the RHRP grant have 103
started rebuilding - by district impact and district (IRM-4, weighted)
Table 6.16: Estimated costs of rebuilding/reconstruction (NPR) - by district 104
(IRM-4, weighted)
Table 6.17: Share of rebuilding costs that the RHRP grant will cover - by district 104
impact and district (IRM-4, weighted)
Table 6.18: Awareness of NPR 100,000 retrofitting grant - by district impact and 105
district (IRM-4, weighted)
Table 6.19: Rating communication with the NRA - by district impact and district 106
(IRM-4, weighted)
Table 7.1: Share of people who say someone in their family got sick due to shelter issues - 108
by district impact and district (IRM-3, IRM-4 household panel, unweighted)
Table 7.2: Share of people who say someone in their family got sick due to shelter 108
issues - by pre-earthquake income, age, disability, gender, widows and
caste (IRM-3, IRM-4 household panel, unweighted)
Table 7.3: Illness prevalence amongst those who got ill (IRM-3, IRM-4 household 110
panel, unweighted)
Table 7,4: Illness prevalence amongst those who got ill - by age (IRM-3, IRM-4 111
household panel, unweighted)
Table 7.5: Share of people reporting psychological effects from the earthquakes - 111
by district impact and district (IRM-3, IRM-4 household panel, unweighted)
Table 7.6: Share of people reporting psychological effects from the earthquakes - 112
by pre-earthquake income, age, disability, gender, widows and caste
(IRM-3, IRM-4 household panel, unweighted)
Table 7,7: Share of people reporting psychological effects from the earthquakes - 112
by housing damage (IRM-3, IRM-4 household panel, unweighted)
Table 7.8: Share of people reporting psychological effects from the earthquakes - 113
by where people were living (IRM-3, IRM-4 household panel, unweighted)
Table 7.9: Share of people reporting psychological effects from the earthquakes among 113
those in temporary or self-constructed shelters - by type of shelter (IRM-3,
IRM-4 household panel, unweighted)
Table 7.10: Type of psychological effect - by district impact and district (IRM-3, IRM-4 114
household panel, unweighted)
Table 8.1: Satisfaction with local political parties since the last monsoon - by district 117
impact and district (IRM-4, weighted)
Table 8.2: Satisfaction with local administrations since the last monsoon - by district 118
impact and district (IRM-4, weighted)
Table 8.3: Satisfaction with local political parties since the last monsoon - by caste, 119
pre-earthquake income and education (IRM-4, weighted)
Table 8.4: Satisfaction with local administration since the last monsoon - by caste, 119
pre-earthquake income and education (IRM-4, weighted)
Table 8.5: Share of people who agree VDC/municipality has distributed aid fairly 120
since the end of the last monsoon - by caste and pre-earthquake income
(IRM-4, weighted)
Table 8.6: Satisfaction with local political parties - by perceptions of whether 120
VDC/municipality has been distributing aid fairly (IRM-4, weighted)
Table 8.7: Satisfaction with how local political parties inform about aid - 121
by perceptions of whether VDC/municipality has been distributing aid
fairly (IRM-4 weighted)
Table 8.8: Satisfaction with local administration - by perceptions of whether 121
VDC/municipality has been distributing aid fairly (IRM-4, weighted)
Table 8.9: Visits from elected officials over time - by district impact and district 122
(IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted)
Table 8.10: Factors determining voting choice - by district impact and district 124
(IRM-4, weighted)
xix


Aid and Recovery in Post-Earthquake Nepal
Table 8.11: Factors determining voting choice - by remoteness (IRM-4, weighted) 124
Table 8.12: Factors determining voting choice - by housing damage (IRM-4, weighted) 125
Table 8.13: Factors determining voting choice - by caste, pre-earthquake income and 125
education (IRM-4, weighted)
Table 8.14: Whether local elections will be free and fair - by district impact and district 126
(IRM-4, weighted)
Table 8.15: Whether local elections will be free and fair - by gender, widows and caste 126
(IRM-4, weighted)
Table 8.16: Why elections will not be free and fair among those who think they will not 127
be free and fair - by district and district impact (IRM-4, weighted)
Table 8.17: Outlook on earthquake reconstruction work as a result of the local elections - 129
by district impact and district (IRM-4, weighted)
Table 8.18: Outlook on earthquake reconstruction work as a result of the local elections - 129
by pre-earthquake income and age (IRM-4, weighted)
Table 8.19: Outlook on earthquake reconstruction work as a result of the local elections - 129
by gender, caste and disability (IRM-4, weighted)
Table 9.1: Proportion feeling very or somewhat unsafe and insecure in their 133
community - by district impact and district (IRM-1, IRM-2, IRM-3, IRM-4
household panel, unweighted)
Table 9.2: Share saying there was a violent incident in their community - by district 134
impact and district (IRM-1, IRM-2, IRM-3, IRM-4 household panel, unweighted)
Table 9.3: Share trusting different groups of people - by district impact and district 136
(IRM-3, IRM-4 household panel, unweighted)
Table 9.4: Share saying people in their community would be very likely to cooperate - by 138
district impact and district (IRM-2, IRM-3, IRM-4 household panel, unweighted)
Table 9.5: Share saying people in their community would be very likely to cooperate - 138
by gender, widows and pre-earthquake income (IRM-2, IRM-3, IRM-4
household panel, unweighted)
Table 9.6: Relations with neighbors after the earthquake - by district impact and
district (IRM-3, IRM-4 household panel, unweighted) 140
Table B.i: Distribution of demographic and socio-economic characteristics - 149
by district impact and district (IRM-4)
Table B.2: Distribution of types of disability - by district impact and district (IRM-4) 151
Table B.3: Rural/urban and remoteness distribution - by district impact and district (IRM-4) 152
Table B.4: Housing damage distribution - by district impact and district (IRM-4) 153
Table B.4: Where are people living now - by district impact and district (IRM-4) 153
XX





Photo: Nayan Pokharel


Photo: Chiran Manandhar
This report provides findings from a large-scale
survey conducted in April 2017, two years on from
the devastating earthquakes that hit Nepal. Much
has changed since the earthquakes. In the first few
months, emergency relief was widely provided, aiming
to meet immediate needs including helping people
find temporary shelter, address food shortages and
ensure disease did not spread. As months went on,
aid changed in volume and form with cash grants
disbursed and livelihoods support provided. Two
years on, with many people still living in temporary
shelter, attention is now squarely focused on getting
people back into permanent housing and rebuilding
other infrastructure.
To what extent have such changes in aid responses
addressed the evolving challenges people face in re-
covering from the disasters? The report documents
the nature of the aid response as of April 2017 and
the degree to which people—of different demographic
groups, suffering from different degrees of impact
from the earthquakes, living in different areas—are
recovering. It looks at a range of issues including
where people are living, the extent to which they are
rebuilding, how their livelihoods are recovering, what
illnesses and trauma they are experiencing and the
coping strategies people are using. It also explores
the environments in which the earthquake-affected
live - how social relations, politics, local economies
and formal and informal institutions are evolving in
1.1 Background
response to the disasters. Such information can help
policy-makers, development practitioners and others
tailor their responses to help speed recovery.
The report is part of the Independent Impacts and
Recovery Monitoring for Accountability in Post-Earth-
quake Nepal (IRM) project which commenced five
weeks after the first earthquake. Using both quanti-
tative surveying 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.
This report provides quantitative findings from the
fourth wave of surveying (referred to as IRM-4). It is
published in parallel with a report outlining the qual-
itative data and a report synthesizing findings.1 The
1 The Asia Foundation and Democracy Resource Center Nepal
(2017). Aid and Recovery in Post-Earthquake Nepal: Independent
Impacts and Recovery Monitoring Nepal Phase 4 - Qualitative
Field Monitoring (April 2017). Kathmandu and Bangkok: The Asia
Foundation; The Asia Foundation (2017). Independent Impacts
and Recovery Monitoring Nepal Phase 4 (April 2017) - Synthesis
Report. Kathmandu and Bangkok: The Asia Foundation.
1


Introduction
Photo: Ishwari Bhattarai
report provides data and analysis on the situation in
April 2017, almost two years on from the first earth-
quake. It compares data collected in April with that
gathered in three previous rounds: IRM-1, conducted
in June 2015; IRM-2, conducted in February-March
2016; and IRM-3 conducted in September 2016.2
Focus areas
The report focuses on a number of areas. For each, it
looks both at the current situation as well as changes
since the earthquakes:
• People’s shelter conditions - where people are
living and progress on reconstruction (Chapter 2);
• The extent to which livelihoods are recovering
and the state of food security and public services
(Chapter 3);
• The coping strategies employed by the affected
and their effectiveness (Chapter 4);
• The nature of the aid response since the end of
the last monsoon season and its fit with needs
(Chapter 5);
• The programs and work of the National Recon-
struction Agency, including the flagship housing
cash grant programs (Chapter 6);
• How living conditions affected health in the past
six months and psychological effects from the
earthquake (Chapter 7);
• How local politics have evolved with a particular
focus on the local elections (Chapter 8);
• Changes in security and social relations in affect-
ed areas (Chapter 9).
The report concludes with a summary of the main
findings and a discussion of some of their implications.
Annexes provide more details on the methodology
employed. The analysis is that of the authors rather
than the funders of IRM.
2 Reports from previous rounds can be accessed at: http://
asiafoundation.org/tag/independent-impacts-and-recovery-
monitoring-nepal /
2


Aid and Recovery in Post-Earthquake Nepal
1.2 Methodology and approach
Sample
The IRM-4 survey involved face-to-face interviews
with 4,854 respondents in 308 wards across 11 earth-
quake-impacted districts which have been studied
since IRM-1 (Map 1.1).3
Map 1.1: Location of surveyed districts (IRM-4)
The eleven IRM survey districts fall into four of the
Post-Disaster Needs Assessment (PDNA) categories
(Table 1.1). Throughout the report, we use these PDNA
classifications when presenting the data. (Severely
hit districts are those deemed most affected; moving
towards the right in the table, districts are less
affected.)
Table 1.1: Districts surveyed (IRM-4)
Severely hit Crisis hit Hit with heavy losses Hit
Ramechhap Okhaldhunga Solukhumbu Syangja
Gorkha Bhaktapur Lamjung
Sindhupalchowk Kathmandu
Nuwakot
Dhading
A full discussion of the methodology is included in Representative data. The data are representative
Annex A. However, two aspects of the approach are of all people in the eleven districts studied. A careful
especially important. sampling strategy—at the Village Development Com-
3


Introduction
mittee (VDC), ward, household and individual levels—
was employed. Stratified random sampling, along with
weighting of the data (discussed below), 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. It should be noted
that the large sample size allows for more accurate
estimates, and that the margins of error are smaller
compared to most surveys, in Nepal and beyond.
From IRM-2 onward, additional households were
sampled in four districts (Sindhupalchowk, Ramech-
hap, Gorkha and Okhaldhunga) to allow for a deeper
assessment of the food (in)security situation. The
Nepal Food Security Monitoring System (NeKSAP)
collects monthly data from local leaders that allows
them to track changes in such insecurity.3 4 To help verify
this, and to see how food insecurity is linked to other
measures of vulnerability, NeKSAP data was used to
select an additional 250 houses in these four districts.5 * * * * *
The margin of error for these four districts is +/- 4%.
These datasets for IRM-1, IRM-2, IRM-3 and IRM-4
are referred to as full datasets.
Tracking changes over time. IRM is set up as a
panel survey - where possible, the same people are
interviewed in each round (referred to as the household
panel dataset). 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 ma-
jority of people interviewed in the IRM-4 survey (4,131
out of the 4,854) had also been interviewed in IRM-2
and 3. A smaller number of these people (1,403) were
also interviewed in IRM-1.6 For some analyses we use
the full datasets from IRM-1, IRM-2, IRM-3 and IRM-
4. For others, we use the household panel datasets.
Analysis
The rich survey data are used in a number of ways
throughout this report.
First, for many analyses we compare the full data of
IRM-1 to IRM-4 at the aggregate level, allowing for an
assessment of changes over time. The IRM-4 survey
was deliberately designed to mirror the previous IRM
instruments, with many of the questions remaining the
same. This allows for direct assessment to be made
of changes over time. Some adjustments were made
between each survey to capture particularly important
events such as the fuel crisis, the cash grant distribu-
tion and, in this round, the upcoming local elections.
The first survey tracked attitudes, perceptions, and
experiences two months after the disaster and changes
since the earthquakes. Most of the IRM-2 questions
recorded information on what had happened between
then and February 2016 when the second survey was
conducted, with the beginning of the 2015 monsoon
period (June 2015) used as the time marker. The
IRM-3 survey, conducted towards the end of the mon-
soon, recorded changes since IRM-2 at a time when
the third official damage assessment was being, or
had recently been, conducted. IRM-4 was conducted
during April 2017 with some important changes oc-
curring since the last round. The reconstruction grant
was increased to NPR 300,000 and the government
announced provision of a NPR 100,000 retrofitting
grant for those whose houses were categorized as par-
tially damaged. The survey was completed two weeks
before the first polling of the first local elections in 20
years. For the new questions, estimates in the study
are based on the IRM-4 dataset alone. Where we use
the full datasets, the data are weighted to ensure they
are representative of the whole population of earth-
quake-affected districts.7
Second, because many people who were interviewed
in IRM-4 were also interviewed in past rounds, we can
assess with more rigor how individuals’perceptions
and experiences have changed over time. Some of the
3 The IRM-1 survey was conducted in 14 districts. Three of these
districts were dropped for IRM-2, IRM-3 and IRM-4. IRM-1 was
conducted before the government’s Post-Disaster Needs Assess-
ment (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) sur-
veyed in IRM-1 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 See http://neksap.org.np/uploaded/resources/Publications-and-
Research/Food-Security-Bulletins/FSB—46—English.pdf
5 The boosting was done as follows. The 1,400 households in the
main sample (350 per district for each of the four districts) were
first classified per NeKSAP into four categories: minimally food
insecure; moderately food insecure; highly food insecure; and
severely food insecure. Following this, 250 households were added
per district in order to create a total food security sample of 600
households per district, with an even representation across all
relevant NeKSAP classifications for the district. The additional
250 households were added using a random sampling method,
based on a list of households corresponding to each NeKSAP
classification within the district. Analysis of this food security data
is presented in Chapter 3.
6 This is primarily because the sampling strategy changed after
IRM-i with three districts dropped and new wards selected in the
remaining 11 districts.
7 See Annex A for a discussion of the weighting strategy.
4


Aid and Recovery in Post-Earthquake Nepal
analyses in the report draw on the sub-sets of the data
that include only those interviewed in all four rounds
or in the past two or three rounds (the household panel
datasets). Because most respondents were interviewed
in IRM-2 to IRM-4, with fewer also interviewed in
IRM-1, we make more use of the IRM-2 to IRM-4
dataset, except where it is particularly important to
examine changes across all four rounds. All results
from the panel datasets are unweighted to best rep-
resent individuals’ responses over time.
Third, many of the analyses and data breakdowns
compare aggregate responses from each of the PDNA
impact categories: severely hit districts; crisis hit
ones; hit with heavy losses districts; and a hit district.
These analyses provide a broad-brush picture of the
differences (and similarities) between districts with
varying degrees of earthquake impact.
Fourth, most of the analyses are also broken down by
individual districts. Each district has experienced the
earthquake, and the aid response, differently. These
granular analyses allow for an exploration of how
districts vary, say, in aid received, in coping strategies
employed, in attitudes towards local leaders. This level
of disaggregation means that, at times, the report
gets into detailed analysis of the situation in specific
districts. We believe the analyses will be useful, in
particular for those working in particular districts.
Fifth, analyses of the data are broken by a host of
demographic and geographic variables. Different
groups of the population (men/women; people of
different caste; people with different incomes; etc.)
will likely have experienced the earthquake in different
ways. Disaggregating analyses by all these demograph-
ic variables allows for a much finer assessment of dif-
fering patterns of impacts and recovery. The analyses
provide information on which groups of people are
more vulnerable who may require particular attention.
Variables
Most of the variables used in the analyses in this report
are self-explanatory. Following are descriptions for
those that may be less clear.
• Caste. Three nominal measures of caste are
included in the study: high caste, low caste and
Janajatis. High caste refers to all castes except
Dalits in both hill and Terai regions (Bahun,
Chhetri, Thakuri, etc.). Low caste refers to Dalits.
Janajati are all other indigenous ethnic groups,
which are generally considered marginalized.
• Income. Respondents in this study are catego-
rized into three levels of pre-earthquake income:
low income, medium income and high income.
The monthly pre-earthquake income of those
in the low income group is up to NPR 9,999;
the monthly incomes of the medium income
group range from NPR 10,000 to NPR 19,999;
the monthly income of the high income group
is above NPR 20,000. In some cases, we look at
differences in outcomes by current income, using
the same categories.
• Disability. Respondents were asked six ques-
tions on disability, drawing on guidance from the
Washington Group on Disability Statistics. Where
respondents said they have a lot of difficulty or
cannot do any one or more of the following, they
are coded as having a disability. (If they mention
having no or some difficulty, then they are coded
as not having a disability.)
1. Seeing, even if wearing glasses;
2. Hearing, even if using a hearing aid;
3. Walking or climbing steps;
4. Remembering or concentrating;
5. Self-care such as washing or dressing;
6. Difficulty communicating.
• Remoteness. Remoteness has three categories
based on how far the ward is from the district
headquarters. If the ward is less than one hour
from the district headquarters, using the quickest
means of transportation, then it is coded as dess
remote.’ If the ward is 1-6 hours from the district
headquarters, it is coded as ‘remote.’ Finally, if
the ward is located more than 6 hours from the
district headquarters, it is coded as ‘more remote.’
Limitations
The survey data presented here are a result of a care-
ful and methodical sampling design. The results are
representative of the full population of the 11 surveyed
districts. The survey was piloted to ensure that re-
spondents understood questions and adjustments were
made where necessary. Lessons from the effectiveness
of the questions in the previous surveys also helped to
improve the IRM-4 instrument. As noted, the large
sample size means that the estimates in the report are
exceptionally accurate, meaning we can have strong
confidence that the findings are true to reality.
However, and as with all surveys, caution should
be taken when interpreting findings. Quantitative
research has both strengths and weaknesses.
First, surveys provide useful information on the
situation of large numbers of people, selected such
that findings can be generalized across the broader
population in affected areas. However, bivariate
results presented in this study do not explain well the
5


Introduction
underlying factors that determine different situations
and attitudes - for example, why people feel safe or
have not received aid.8
Second, information provided throughout the report
is based on self-reported accounts. Results related
to factual events may not have been captured well
by the survey. For instance, many may not have
full knowledge of the situation (e.g. who provided
aid or whether politicians had visited their wards
or the number of incidents of crime). Others may
have incentives to over- or under-report the level
of impact they experienced, whether or not they
received aid, and so on. While results on average still
tend to represent the general perception among the
population, it is important to bear in mind that these
are self-reported accounts.
Third, some questions, such as whether violence has
occurred are sensitive and some may prefer not to
answer them.
The IRM-4 synthesis report, published separately,
combines information from both the quantitative
survey and the in-depth qualitative fieldwork. This
multi-method approach allows for a triangulation of
findings and a deeper exploration of causal relation-
ships - i.e. what is driving recovery.
8 Throughout this report, we present correlations between outcome
variables and factors that may be associated with them (for
example, whether people received aid and the extent to which their
house was damaged by the earthquakes). But even where we find
close correlations, this does not mean that one causes the other.
6


Aid and Recovery in Post-Earthquake Nepal
Photo: Ishwari Bhattarai
Two years on from the earthquakes, many people in
affected areas were still living in temporary shelters.
This chapter explores where people were living in
April 2017. It analyzes levels of improvement in shelter
and housing as well as causes of delays. It looks at
who has moved from temporary shelter back into
permanent housing, the quality of temporary shelters,
preparedness for adverse weather, and progress in
housing reconstruction. Throughout, it analyzes which
groups are doing better and which are struggling.
Key Findings
Where people are living
• There has been some progress in people moving
from temporary shelters back into their homes.
Almost three-quarters in affected areas lived in
their own homes in April 2017 compared to 60%
in the early weeks after the disaster. However,
almost two-in-three people in severely hit dis-
tricts remained in temporary shelters two years
after the quakes and there has been relatively
little movement since IRM-3 was conducted in
September 2016.
• People in more remote areas are far more likely
to continue to live in shelters (45% remain in
shelters) than those in less remote areas (6%).
• Almost all of the people in temporary shelters
had seen their house completely destroyed by
the earthquakes. Those whose house suffered less
damage are much less likely to remain in shelters.
Almost half of those whose house was destroyed
were still in shelters in April.
• People who live in less-affected districts whose
house was destroyed or suffered major damage
are much more likely to have moved home than
those who live in severely hit districts.
• Those marginalized before the earthquakes (those
with a low income, low education, the disabled)
are much more likely to remain in temporary
shelters.
Movements from shelter to houses
• Among those who were living in temporary shel-
ters in IRM-2, nearly 34% had moved to their
homes in IRM-4, with the rest continuing to live
in shelters. There has been the least movement
home in Sindhupalchowk, Dhading and Nuwakot.
• People in less remote areas are more likely than
others to have moved back home. People in
severely hit districts are substantially less likely
to move to their own houses than those in lesser
affected districts who were previously in shelters.
7


Shelter
Quality of temporary shelters
• Almost everyone in temporary shelters now lives
in shelters that use CGI. A small amount of people
(2% of those in shelters) have been living in cow
sheds since the earthquake.
• In IRM-4, people in less remote areas were more
likely to be in shelters made out of CGI sheets;
shelters found in remote and more remote areas
were more likely to be either non-CGI shelters
or those that combine CGI sheets with wood or
bamboo.
Preparedness for adverse weather
• The majority of respondents in IRM-4 (70%) said
that they were able to fix their shelters sufficiently
or completely in preparation for winter. However,
17% failed to repair their shelters completely and
11% failed to repair sufficiently.
• More people in severely hit districts were unable
to repair sufficiently or completely in preparation
for the adverse weather.
• Individuals of low caste or low income, and those
with disabilities, were less likely to repair suffi-
ciently or completely.
Rebuilding and reconstruction
• The majority of people whose house was com-
pletely destroyed or suffered major damage have
not started to rebuild. Those whose house was
substantially impacted are less likely to have
started rebuilding in districts that were more
affected. People have been most likely to start
rebuilding in Solukhumbu and Syangja.
• People of low caste or low income are less likely
than others to have started rebuilding. Those
whose income level has declined since the earth-
quake are also far less likely to have started
rebuilding.
• When asked what has prevented rebuilding, the
top two reasons cited were not having enough
money and waiting for government cash grants.
• People in more affected districts and more re-
mote areas are more likely to say that they have
not completed rebuilding or built a new house
because they are waiting for government housing
grants.
Photo: Ishwari Bhattarai
8


Aid and Recovery in Post-Earthquake Nepal
2.1 Where are people living?
There has been limited progress since the earthquake
in people moving from temporary shelters back into
their homes. Almost three-quarters of people in earth-
quake-affected areas now live in their own homes
compared to 60% in the immediate aftermath of the
earthquakes (Figure 2.1). Twenty-four percent now live
in self-constructed shelters compared to 33% in IRM-1.
Figure 2.1: Where people were/are living (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) | Sep 2016 (IRM-3)
Feb-Mar 2016 (IRM-2) H Apr 2017 (IRM-4)
In more affected districts, a much larger share of the
population is still living in temporary shelters. Sixty-
two percent of people in the severely hit districts, those
most affected by the earthquakes, are currently living
in temporary shelters. In contrast, only 5% in crisis
hit districts and 2% each in hit with heavy losses and
hit districts are still in temporary shelters (Table 2.1).
Sindhupalchowk has the highest proportion of people
living in temporary shelters either on their own land
(76%) or on other people’s land (8%).9
Table 2.1: Where people are living now - by district impact and district (IRM-4, weighted)
Own house Neighbor’s house Self- constructed shelter on own land Self- constructed shelter on other people’s land Self- constructed shelter on public land Rent
Severely hit 37% 1% 59% 3% 0% 0%
Dhading 34% 1% 64% 2% 0% 0%
Gorkha 55% 1% 40% 4% 0% 1%
Nuwakot 37% 0% 61% 1% 1% 0%
Ramechhap 45% 0% 52% 2% 1% 0%
9 In some of the tables and figures in this report, numbers do not
add up to 100% because of rounding errors.
9


Shelter
Own house Neighbor’s house Self- constructed shelter on own land Self- constructed shelter on other people’s land Self- constructed shelter on public land Rent
Sindhupalchowk 16% 0% 76% 8% 0% 0%
Crisis hit 91% 0% 4% 1% 0% 3%
Bhaktapur 81% 3% 10% 1% 2% 3%
Kathmandu 93% 0% 2o/o 1% 0% 3%
Okhaldhunga 75% 2% 21% 1% 0% 2%
Hit with heavy losses 97% 0% 2% 0% 0% 1%
Solukhumbu 95% 0% 5% 0% 0% 0%
Lamjung 97% 0% 1% 0% 0% 2%
Hit 95% 0% 2% 0% 0% 2%
Syangja 95% 0% 2o/o 0% 0% 2%
A far larger share of the population in more remote
areas is still living in temporary shelters compared to
people in less remote areas (Figure 2.2).10 Forty-five
percent of people in more remote areas are still in
shelters compared to 37% in remote areas and just
6% in less remote areas.
Figure 2.2: Where people are living - by remoteness (IRM-4, weighted)
Less remote
Remote
| More remote
Earthquake impact and temporary shelter
Sixty-two percent of people in affected districts report
that their house was either completely destroyed or
suffered major damage from the earthquakes. More
than 90% of these people are from severely hit or crisis
hit districts (Figure 2.3).
10 Remoteness is categorized according to the time it takes for
individuals to reach the district headquarters from their homes.
Places that are 1 hour or less from district headquarters by the
fastest means of travel are categorized as less remote, 3-6 hours
from district headquarters are remote areas, and those more than
6 hours away are categorized as more remote areas.
10


Aid and Recovery in Post-Earthquake Nepal
Figure 2.3: Share of people whose house
was completely destroyed or suffered major
damage - by district impact (IRM-4, weighted)
4%
Severely hit H Hit with heavy losses
Crisis hit H Hit
*1 % rounding error
Almost half of households whose house was completely
destroyed continue to live in temporary shelters
(Figure 2.4). According to self-reported levels of
housing damage, 44% of people whose house was
completely destroyed are still living in shelters,
compared to 5% whose house was badly damaged.
Almost no-one whose house suffered minor or no
damage was living in a shelter in April 2017.
or destroyed in lesser affected districts have been
much more likely than those in severely hit districts
to move from temporary shelters into their homes.
Amongst the severely hit districts, those in Gorkha
and Ramechhap whose house was destroyed or badly
damaged have been more likely to move home than in
other severely hit districts, especially Sindhupalchowk.
Table 2.2: Share of people with a completely
destroyed or majorly damaged house living
in shelters - by district impact and district
(IRM-4, weighted)
Proportion in temporary shelter
Severely hit 66%
Dhading 70%
Gorkha 47%
Nuwakot 65%
Ramechhap 57%
Sindhupalchowk 86%
Crisis hit 12%
Bhaktapur 25%
Kathmandu 7%
Okhaldhunga 30%
Hit with heavy losses 4%
Solukhumbu 5%
Lamjung 3%
Hit 8%
Syangja 8%
Figure 2.4: Share of people living in temporary
shelters (on own, public or other’s land) -
by level of house damage (IRM-4, weighted)
Of those whose house was completely destroyed or
suffered major damage, 66% in severely hit districts
are still living in shelters (Table 2.2). More than 85%
of the people in other impact categories report living
in their own houses (not shown in the table). The data
show that those whose house was badly damaged
Who is still living in temporary
shelters?
As in previous rounds of IRM, those who were mar-
ginalized before the earthquakes are far more likely to
remain in temporary shelters than others (Table 2.3).
Individuals whose income was low before the earth-
quake (44%) are more likely to still live in shelters than
those whose income was of medium (20%) or high
(10%) levels.11 People with no education are nearly 30
percentage points more likely to be in shelters than
those with the highest level of education (master’s
degree or above). There are also notable differences
between caste groups. Individuals in the high caste
group are substantively less likely than others to still
live in shelters. High caste individuals are 8 percent-
age points less likely than low caste and 9 percentage
points less likely than Janajatis to still live in tempo-
rary shelters.
11 Low income people are those with a monthly income of less than
NPR 10,000. Medium income: NPR 10,000-19,999. High income:
NPR 20,000 or more per month.
11


Shelter
These differences are not a product of more mar-
ginalized groups suffering higher levels of housing
damage. When looking only at those whose house was
completely destroyed or badly damaged, marginalized
groups are much less likely than others whose houses
suffered similar impacts to move from temporary shel-
ters. While 32% of high caste people whose house was
destroyed/badly damaged are in shelters, the figure
is 46% for low caste people. Income and education
levels are also strong predictors of whether or not
people with severe damage to their houses are still
living in shelters. Those with a low pre-earthquake
income whose houses suffered either major damage or
complete destruction are more than twice as likely as
similar individuals in the high income group to still be
living in shelters. When disaggregating by education
level, the probability of people whose house suffered
significant damage still living in shelters does not
decrease linearly with increasing level of education
but people with a master’s degree or above are nearly
three time less likely to be living in shelters compared
to other groups on average.
Table 2.3: Share of people living in temporary shelters (on own, public or other’s land) -
by caste, pre-earthquake income and education (IRM-4, weighted)
Group Proportion in temporary shelter Proportion in temporary shelter (of those whose house was destroyed or badly damaged)
High caste 18% 32%
Caste Janajati 27% 40%
Low caste 26% 46%
Pre-earthquake income Low 44% 54%
Medium 20% 32%
High 10% 21%
Illiterate 37% 49%
Literate but no education 25% 38%
Primary level 27% 38%
Lower secondary level 18% 30%
Education Secondary level 10% 19%
SLC Pass 9% 20%
+2/intermediate pass 17% 38%
Bachelor pass 16% 39%
Master and above 4% 11%
Those with a disability are also more likely than
others to still live in temporary shelters: 38% versus
23% (Table 2.4). While there is not much difference
by gender, a slightly higher proportion of widows re-
port living in shelters in IRM-4 compared to others.
According to the survey, approximately 5.4% of the
affected population are widows and they are 3 percent-
age points more likely than others to live in temporary
shelters. Findings are similar when only looking at
those whose house was destroyed or badly damaged.
Table 2.4: Share of people living in temporary shelters (on own, public or other’s land) -
by gender, widows and disability (IRM-4, weighted)
Group Proportion in temporary shelter Proportion in temporary shelter (of those whose house was destroyed or badly damaged)
Female 24% 37%
Gender Male 23% 38%
Widows 27% 40%
Il 10 0 r"\ 111+\ / No disability 23% 37%
uisaoiiny Disability 38% 52%
12


Aid and Recovery in Post-Earthquake Nepal
2.2 Movements from shelter to houses
Because IRM interviews the same people in each
round of surveying, we can identify how individuals’
housing situation has changed over time. The IRM-2/
IRM-3/IRM-4 panel dataset contains information for
4,131 respondents who were interviewed in all three
rounds. Nearly half of these respondents (unweighted)
were living in some form of temporary shelter in IRM-
2. Among those who were living in temporary shelters
in IRM-2, nearly 34% had moved to their homes in
IRM-4, whereas 65% continue to live in shelters.
Among those living in shelters in IRM-2, a greater
share of people in less remote and lower impact
districts moved to their houses. As shown in Figure 2.5,
people in less remote areas are 10 percentage points
more likely to have moved to their homes compared
to those in remote areas and 5 percentage points
more likely than those in more remote areas. People
in severely hit districts are twice less likely to move to
their own houses than those in hit with heavy losses
or hit districts.
Figure 2.5: Share of people who were living in shelter in IRM-2 who moved to their own house in IRM-4
(IRM-2, IRM-4, household panel, unweighted)
Who has continued to live in shelters since IRM-2?
Among the subset of people living in shelters in IRM-
2, nearly 65% continue to live in shelters in IRM-4
(Table 2.5). The houses of 99% of those who continue
to live in shelter were completely destroyed or
suffered major damage (not shown in the table). When
disaggregating by impact categories, the majority of
such people living in severely hit districts continue to
live in shelters in IRM-4.
Movements out of shelters have been least common
in Sindhupalchowk (where 85% of those in shelters at
IRM-2 remain in temporary shelters), Nuwakot (65%)
and Dhading (72%).12 In lesser affected districts, most of
those who were in shelters in IRM-2 have moved home.
Those in remote and more remote areas who were
in shelters in IRM-2 have been less likely to move
home. Among people living in shelters in IRM-2,
remote areas in IRM-4 have 10 percentage points more
people, and more remote areas have 5 percentage
points more people who continue to live in shelters
compared to those in less remote regions.
12 With caution that these are non-weighted estimates
13


Shelter
Table 2.5: Share of people living in shelters in
IRM-2 who continue to live in shelters in IRM-
4 - by district impact and district (IRM-2, IRM-4,
household panel, unweighted)
Live elsewhere (IRM-4) Continue to live in shelter (IRM-4)
Severely hit 31% 69%
Dhading 28% 72%
Gorkha 40% 60%
Nuwakot 35% 65%
Ramechhap 40% 60%
Sindhupalchowk 15% 85%
Crisis hit 57% 43%
Bhaktapur 45% 55%
Kathmandu 65% 35%
Okhaldhunga 60% 40%
Hit with heavy losses 89% 11%
Lamjung 96% 4%
Solukhumbu 81% 19%
Hit 67% 33%
Syangja 67% 33%
All districts 35% 65%
Less remote 43% 57%
Remote 33% 67%
More remote 38% 62%
Low caste people who were in shelters in IRM-2 are
more likely than others, especially those of high caste,
to continue to live in shelters (Table 2.6). Income
is another important factor. Individuals in the low
income group are 11 percentage points more likely
than the high income group and 7 percentage points
more likely that the medium income group to continue
to live in shelters. Education levels and gender are not
good predictors of who remains in shelters. However,
widows are 4 percentage points more likely than
others, and people with disabilities are 5 percentage
points more likely than those without any disabilities,
to continue living in shelter in IRM-4.
In short, the analyses—both of where people are living
currently, of where those whose house was severely
damaged are living now, and of where people are living
now who were in shelters during IRM-2—all point
to the same conclusions: the marginalized are much
more likely to remain in shelters and have found it
much harder to move home.
Table 2.6: Share of people living in shelters in IRM-2 who continue to live in shelters
in IRM-4 - by caste, pre-earthquake income, education, gender, widows and disability
(IRM-2, IRM-3, IRM-4, household panel, unweighted)
Live elsewhere (IRM-4) Continue to live in shelter (IRM-4)
High caste 40% 60%
Caste Janajati 33% 67%
Low caste 30% 70%
Low 31% 69%
Pre-earthquake Medium 38% 62%
income High 42% 58%
Don’t know/refused 52% 48%
Illiterate 32% 68%
Literate but no education 35% 65%
Primary level 29% 71%
Lower secondary level 39% 61%
Education Secondary Level 54% 46%
SLC pass 44% 56%
+2/lntermediate pass 36% 64%
Bachelor pass 50% 50%
Master and Above 25% 75%
Female 35% 65%
Gender Male 35% 65%
Widows 31% 69%
n 10 0111+\ / No disability 35% 65%
uisaoimy Disability 30% 70%
14


Aid and Recovery in Post-Earthquake Nepal
2.3 Quality of temporary shelters
As with previous rounds of IRM, almost all of those
living in shelters live in either CGI shelters or shelters
made from a combination of CGI and wood or bamboo
(Figure 2.6). There has been a decline in the number
of people in shelters that use tarpaulins or that are
primarily built from bamboo.
Figure 2.6: Share of people living in different types of shelters (IRM-2, IRM-3, IRM-4, weighted)
wood/bamboo
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
When disaggregating shelter types by geographical
factors, variation across remoteness is most striking.
In IRM-4, those in shelters in less remote areas are
more likely to be in shelters made out of CGI sheets,
but shelters found in remote and more remote areas
are more likely to be either non-CGI shelters or those
that combine CGI sheets with wood or bamboo. The
share of people using CGI sheets only in less remote
areas is almost twice that of those in shelters in more
remote areas and 21 percentage points higher than in
remote areas (Table 2.7).
A relatively smaller share of people in shelters in se-
verely hit districts are in CGI only shelters compared to
those in lesser affected districts. Individuals in shelters
in Dhading, Gorkha, Ramechhap, Okhaldhunga and
Lamjung are the least likely to be in CGI only shelters.
Okhaldhunga and Solukhumbu have the highest pro-
portions of people (21% and 12% of those remaining
in shelters, respectively) who are living live in bamboo
shelters, and Syanja has the highest share of people in
shelters who are living in cowsheds (22%).
Table 2.7: Share of people living in different types of shelters -
by district impact, district and remoteness (IRM-4, weighted)
CGI shelter Wood shelter Bamboo shelter Cowshed Tent CGI + tarp CGI + wood/ bamboo Cemented house Mud house
Severely hit 39% 4% 0% 1% 0% 0% 53% 0% 1%
Dhading 30% 1% 0% 1% 0% 0% 65% 0% 3%
Gorkha 38% 12% 1% 3% 1% 1% 42% 0% 2%
Nuwakot 53% 3% 0% 0% 0% 0% 43% 0% 0%
Ramechhap 11% 0% 1% 4% 0% 1% 83% 0% 0%
Sindhupalchowk 50% 5% 0% 0% 0% 0% 43% 0% 0%
Crisis hit 69% 3% 7% 4% 1% 1% 9% 5% 1%
Bhaktapur 80% 2% 7% 0% 2% 0% 4% 2% 2%
Kathmandu 92% 0% 0% 0% 0% 0% 0% 8% 0%
15


Shelter
CGI shelter Wood shelter Bamboo shelter Cowshed Tent CGI + tarp CGI + wood/ bamboo Cemented house Mud house
Okhaldhunga 12% 11% 21% 17% 1% 5% 31% 0% 2%
Hit with heavy losses 66% 0% 9% 9% 0% 0% 16% 0% 0%
Lamjung 33% 0% 0% 0% 0% 0% 67% 0% 0%
Solukhumbu 76% 0% 12% 12% 0% 0% 0% 0% 0%
Hit 66% 11% 0% 22% 0% 0% 0% 0% 0%
Syangja 66% 11% 0% 22% 0% 0% 0% 0% 0%
All districts 43% 4% 1% 2% 0% 1% 47% 1% 1%
Less remote 62% 2% 2% 0% 1% 0% 27% 5% 1%
Remote 41% 4% 1% 2% 0% 0% 51% 0% 1%
More remote 34% 11% 2% 3% 1% 1% 46% 0% 2%
Income level is strongly associated with the type of
shelter people live in. As shown in Table 2.8, high
income individuals are 4 percentage points more
likely than those in the medium income group and 12
percentage points more likely than those with a low
income to live in CGI only shelters. The differences are
less distinct by caste, gender and widows, but people
in shelters with a disability are 5 percentage points
less likely to be living in CGI only shelters than others.
Table 2.8: Share of people living in different types of shelters - by caste, pre-earthquake income,
education, gender, widows and disability (IRM-4, weighted)
CGI shelter Wood shelter Bamboo shelter Cowshed Tent CGI + tarp â– 0 0 0 + -Q O -Q Cemented house Mud house
High caste 43% 4% 1% 3% 0% 0% 45% 2% 1%
Caste Janajati 44% 4% 1% 1% 0% 0% 47% 0% 1%
Low caste 40% 5% 0% 0% 0% 2% 52% 0% 1%
Pre-earthquake income Low 39% 4% 1% 2% 1% 1% 51% 0% 1%
Medium 47% 4% 1% 1% 0% 0% 44% 0% 2%
High 51% 1% 0% 1% 0% 0% 41% 4% 2%
Illiterate 43% 4% 1% 2% 0% 1% 48% 0% 1%
Literate but no education 36% 5% 2% 3% 0% 0% 53% 1% 1%
Primary level 46% 5% 1% 1% 0% 0% 45% 0% 1%
Lower secondary level 45% 2% 1% 1% 1% 1% 50% 0% 0%
Education Secondary level 40% 6% 1% 2% 2% 0% 49% 0% 0%
SLC pass 49% 5% 0% 1% 0% 1% 41% 0% 3%
+2/lntermediate pass 66% 5% 0% 0% 1% 0% 25% 0% 2%
Bachelor pass 55% 5% 0% 0% 4% 0% 9% 26% 0%
Master and above 20% 0% 0% 0% 0% 0% 80% 0% 0%
Female 42% 5% 1% 2% 0% 1% 48% 0% 1%
Gender Male 45% 4% 2% 2% 1% 0% 45% 1% 1%
Widows 45% 3% 0% 2% 0% 1% 47% 1% 1%
Disability No disability 44% 4% 1% 2% 0% 0% 47% 1% 1%
Disability 39% 4% 2% 3% 1% 1% 48% 0% 2%
16


Aid and Recovery in Post-Earthquake Nepal
2.4 Preparedness for adverse weather
When asked if respondents were able to prepare their
shelters for winter weather, the majority (70%) in
IRM-4 said that they were able to fix them sufficient-
ly or completely. However, 17% failed to repair their
shelters completely and 11% failed to repair them
sufficiently. When comparing with people’s prepar-
edness in the past, as shown in Figure 2.7, a relatively
higher share of people were able to completely repair
their shelters in IRM-4 (14%) compared to IRM-3
(6%) or IRM-2 (3%). The share of people who failed
to repair sufficiently or completely has not changed
since IRM-3.
Figure 2.7: Share of people preparing their shelters for winter (IRM-2, IRM-4)/
monsoon (IRM-3) (IRM-2, IRM-3, IRM-4, weighted)
sufficient for winter sufficient for winter
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3)
| Apr 2017 (IRM-4)
*7 % in IRM-2 and IRM-4 mentioned that their shelter did not need any repair
Photo: Chiran Manandhar
17


Shelter
In IRM-4, more people in severely hit districts were
only able to make inadequate or no repairs in prepa-
ration for adverse weather than was the case in lesser
affected districts (Table 2.9). Thirty percent of people
in severely hit districts made inadequate or no repairs,
compared to 12% in crisis hit districts. Although this
figure is high in the other two categories, there are less
than 1% respondents in each, thus making it difficult
to make any meaningful comparison. Responses on
repairing shelters do not vary clearly by remoteness.
Three-quarters of people in less remote areas were
able to sufficiently or completely fixed their shelters,
while 70% in remote and 75% in more remote were
able to do the same.
Table 2.9: Share of people preparing their shelters for winter - by district impact,
district and remoteness (IRM-4, weighted)
Sufficient repair Completely fixed the house No repair Inadequate repair Improving the shelter was not necessary
Severely hit 53% 16% 19% 11% 1%
Dhading 25% 40% 10% 25% 0%
Gorkha 48% 20% 20% 10% 3%
Nuwakot 65% 3% 28% 4% 0%
Ramechhap 50% 5% 26% 19% 0%
Sindhupalchowk 73% 6% 17% 3% 1%
Crisis hit 78% 5% 8% 4% 4%
Bhaktapur 73% 11% 4% 11% 0%
Kathmandu 83% 0% 8% 0% 8%
Okhaldhunga 75% 9% 11% 4% 2%
Hit with heavy losses* 80% 8% 0% 12% 0%
Lamjung* 33% 33% 0% 33% 0%
Solukhumbu* 94% 0% 0% 6% 0%
Hit* 22% 0% 22% 55% 0%
Syangja* 22% 0% 22% 55% 0%
All districts 56% 14% 17% 11% 1%
Less remote 70% 5% 14% 10% 2%
Remote 53% 17% 18% 11% 1%
More remote 62% 13% 17% 7% 0%
★Less than 1% respondents
Figure 2.8: Share of people preparing their shelters for winter - by caste (IRM-4, weighted)
High caste | Janajati H Low caste
18


Aid and Recovery in Post-Earthquake Nepal
Individuals in the low caste group were less likely to
repair sufficiently or completely compared to Janajatis
or those in the high caste group (Figure 2.9). Low caste
people were more likely to report inadequate or no
repair (37%) compared to Janajatis (27%) or those of
high caste (29%).
Among those living in shelters, individuals with a low
pre-earthquake income were less likely to prepare their
shelters for the winter compared to others (Figure 2.9).
Slightly more people in the medium and high income
groups said that they were able to sufficiently or com-
pletely repair their shelters for winter with a relatively
larger share of people in the low income group saying
that they had made either inadequate or no repairs
(30%) compared to medium income individuals (25%)
and high income people (23%).
People with disabilities have also faced greater difficulty
repairing their shelters (Figure 2.10). Seventy-one
percent of people without any disability said that
they had been able to sufficiently or completely repair
the shelters compared to only 54% of people with
disabilities.
Figure 2.9: Share of people preparing their shelters for winter - by pre-earthquake income
(IRM-4, weighted)
Low income H Medium income H High income
Figure 2.10: Share of people preparing their shelters for winter - by disability
(IRM-4, weighted)
No disability Disability
19


Shelter
Photo: Ishwari Bhattarai
2.5 Rebuilding and reconstruction
Table 2.10: Proportion whose house was destroyed
or suffered major damage who have done nothing
to rebuild their damaged house - by district impact
and district (IRM-4, weighted)
Proportion who have not started rebuilding
Severely hit 62%
Dhading 69%
Gorkha 49%
Nuwakot 58%
Ramechhap 61%
Sindhupalchowk 70%
Crisis hit 55%
Bhaktapur 53%
Kathmandu 55%
Okhaldhunga 60%
Hit with heavy losses 42%
Solukhumbu 34%
Lamjung 52%
Hit 35%
Syangja 35%
All 56%
Less remote 52%
Remote 59%
More remote 48%
The majority of people whose house was impacted
by the earthquakes have not started rebuilding their
houses. When asked if they have started rebuild-
ing, 56% of those whose house suffered complete
destruction or major damage reported that they had
not done anything to rebuild. Those in more affected
districts are less likely to have started rebuilding
(Table 2.10). Among those whose houses were com-
pletely destroyed or suffered major damage, 62% in
severely hit districts, 55% in crisis hit, 42% in hit with
heavy losses and 34% in hit districts have done nothing
to rebuild. People in Sindhupalchowk, Ramechhap
and Okhaldhunga are the least likely to have started
rebuilding, while a larger share of people whose house
was destroyed or suffered major damage have started
rebuilding in Solukhumbu and Syangja.
In contrast to earlier findings, people in more remote
areas are not less likely to have started rebuilding. For-
ty-eight percent have not started rebuilding, compared
to 59% in remote areas and 52% in less remote areas.
When did people start rebuilding?
Respondents in IRM-4 were asked when they started
rebuilding their house or building a new house. Overall
32% responded to this question, out of whom only 12%
started the rebuilding process before the first monsoon
immediately after the earthquake (Figure 2.11).
The share of people who started rebuilding process
increases over time, with 16% saying that they started
rebuilding during the first monsoon and 21% after the
first monsoon before the first winter. The number of
people then dropped until after the second winter,
when 17% of people started to rebuild.
20


Aid and Recovery in Post-Earthquake Nepal
Figure 2.11: Time when respondents started to rebuild their house or build
a new house of those who have started to rebuild (IRM-4, weighted)
Before the first monsoon
During the first monsoon
After the first monsoon,
before the first winter
During the first winter
After the first winter,
before the second monsoon
During the second monsoon
After the second monsoon
and before the second winter
During the second winter
After the second winter
Who has still to start rebuilding?
People of low caste or low income are less likely than
others to have started rebuilding (Figure 2.12). Those
of low caste are 9 percentage points more likely than
high caste people, and 4 points more likely than
Janajatis, to have not started rebuilding. Low income
individuals are 1 percentage point more likely than
those with a medium income and 4 points more likely
than those with a high income to have not started
rebuilding. Differences by gender and disability are
negligible, but widows are 4 percentage points less
likely than others to have started rebuilding their
damaged or destroyed house.
Figure 2.12: Proportion who have not done anything to rebuild their damaged house - by caste,
pre-earthquake income, education, gender, widows and disability (IRM-4, weighted)
21


Shelter
Change in income and delay in rebuilding
Those whose income level has declined since the declined since the earthquake are 14 percentage points
earthquake are far less likely to have started rebuilding, more likely than those whose income has improved to
As shown in Figure 2.13, those whose income has say they have done nothing to rebuild their houses.
Figure 2.13: Change in income since the earthquake and delay in rebuilding house (IRM-4, weighted)
Improved
Lowered
No change
What has prevented people from rebuilding their houses?
When asked about the reasons why they have not
starting rebuilding, people overwhelmingly mentioned
not having enough money (93%) - Figure 2.14.13
Compared to the IRM-3 survey, conducted six months
earlier, this is an increase of 4%.14 The next most
common reason according to respondents in IRM-
4 was that they were waiting for government cash
grants (49%). The share of people citing this reason
has declined from 66% in IRM-3.
Figure 2.14: Reasons for stopping repairing or not building a house (IRM-4, weighted)
13 Respondents could give multiple reasons why they have not
started rebuilding, hence percentages do not add up to 100%.
14 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 3: September 2016. Quantitative Report. Kathman-
du and Bangkok: The Asia Foundation, p. 24.
22


Aid and Recovery in Post-Earthquake Nepal
Waiting for government cash grants before rebuilding
People in more affected districts in IRM-4 are more
likely to say that they have not completed the rebuild-
ing process or built a new house because they are
waiting for government housing grants (Table 2.11).15
Compared to only 24% in hit districts, more than 50%
of people in higher impact districts in IRM-4 say they
have not yet rebuilt because they are still waiting for
government cash grants. When comparing these re-
sults with responses in IRM-3, there is a substantial
decline in the share of people in severely hit districts
who cite waiting for government cash grants as a rea-
son for not having rebuilt, from 84% to 51%, but an
increase in hit with heavy losses (by 9 points) and hit
districts (by 20 points).
The share of people still to complete rebuilding their
houses in IRM-4 who say they are waiting for cash
grants from the government also correlates with
remoteness. Compared to 47% of people in less remote
regions, 50% in remote regions and 58% in more
remote regions say they are waiting for government
cash grants to rebuild their houses. Comparing with
IRM-3, this decline is higher in remote regions (by 23
points) and more remote regions (19 points) than in
less remote areas (3 points).
Table 2.11: Proportion who have stopped the rebuilding process or not built a house because waiting
for government cash grants - by district impact and remoteness (IRM-3, IRM-4, weighted)
Still waiting for government cash grant IRM-4 Still waiting for government cash grant IRM-3
Severely hit 51% 84%
District impact Crisis hit 50% 51%
Hit with heavy losses 54% 45%
Hit 24% 4%
Less remote 47% 50%
Remoteness Remote 50% 73%
More remote 58% 77%
Low caste people are 9 percentage points more likely
than high caste individuals to wait for government
cash grant to complete rebuilding their house or to
start building a new house. Men are 5 percentage
points more likely than women to wait for government
cash grants. Widows are 15 percentage points more
likely than others to wait for government cash grants
(Table 2.12)
Table 2.12: Reasons for stopping the rebuilding process or not building a house - by caste,
pre-earthquake income, gender, widows, disability and housing damage (IRM-4, weighted)
Did not have enough money Still waiting for government cash grant
High caste 91% 47%
Caste Janajati 94% 50%
Low caste 97% 58%
Pre-earthquake income Low income 94% 47%
Medium income 93% 52%
High income 92% 52%
Gender Female 93% 47%
Male 93% 52%
Widows
Widows 95% 63%
15 See Chapter 6 for analysis of the government cash grant programs.
23


Shelter
Did not have enough money Still waiting for government cash grant
Disability No disability 93% 49%
Disability 92% 47%
Completely destroyed 94% 53%
Housing damage Badly damaged (needs major repair to live in) 92% 49%
Habitable (but needs minor repair) 89% 24%
Not damaged 100% 3%
Those who suffered a higher degree of damage to their
houses are more likely to wait for the government cash
grant before completing the rebuilding process or
building a new house. Over half of people whose house
was completely destroyed say they are waiting for gov-
ernment cash grants, compared to 49% whose house
saw major damage, and 24% with minor damage.
24


Photo: Chiran Manandhar

People’s housing situation is just one measure of their
recovery. This chapter explores three other elements
of recovery: livelihoods and income sources, food
need and consumption, and access to and satisfaction
with public services. By comparing the progress in the
last eight months with that of the past, the chapter
highlights areas where more attention is needed.
Key Findings
Recovery of livelihoods
• Over time, there has been a large drop in the num-
ber of people generating income mainly through
farming. In contrast, far more people are generat-
ing income through their own businesses or daily
wage work than in the past and remittances have
become more important.
• People in severely hit and crisis hit districts, along
with those living in shelters, are more likely to
have seen a decline in income compared to those
in lower impact districts. Declining income has
been particularly widespread in Sindhupalchowk
and Gorkha districts. Income recovery in more re-
mote areas is lagging behind that in other regions.
• People with higher pre-earthquake income and
those who are more educated are more likely to have
seen their income increase since the earthquakes.
Food
• Far fewer people are saying the provision of food
is a priority need than in previous IRM surveys.
However, food continues to remain an acute need
in some areas and for certain groups of people.
Those in severely hit districts are much more like-
ly than others to say they need food. Food demand
is also much higher in more remote areas.
• Stated need for food is substantially higher among
people who are low caste, whose income is low,
and for widows and people with a disability.
• Thirty-four percent now say their food consump-
tion has increased over the past year compared
to 31% in February 2016.
• Those in severely hit districts are substantially
more likely to report decreases in consumption.
Districts with the most frequent reports of de-
creases in food consumption are Okhaldhunga
(13% of people), Sindhupalchowk (12%), Nuwakot
(13%) and Dhading (12%).
Public services
• Nine percentage points fewer people say they have
access to drinking water in April 2017 compared
to September 2016. Access to clean drinking
water has declined in severely hit and crisis hit
25


Shelter
districts. The two districts where problems with
clean drinking water problem seem acute are
Gorkha and Nuwakot.
• People of lower caste are generally more likely to
be dissatisfied with most public services, although
not by a large margin. A notable exception is
drinking water, with 31% of low caste people
expressing dissatisfaction compared to 25% of
high caste people and 22% of Janajatis. This
reflects ongoing problems many low caste people
have in accessing public water supplies.
• Those with a low income are more likely to express
dissatisfaction with electricity, medical facilities
and schools, but less likely to be dissatisfied with
drinking water and roads. Those with a high
income are the most likely to be dissatisfied with
the quality of roads.
3.1 Recovery of livelihoods
What are people’s income sources and are
people changing livelihoods?
Since the earthquakes, there has been a decline in
the number of people who generate income through
farming and an increase in the number generating
income through their own business, daily wage work or
remittances. When respondents were asked whether
they had changed their livelihood in the previous three
months, only 1% in IRM-2 said they had done so while
2% in IRM-3 and IRM-4 said they had. However, while
these numbers are low, the data also reveal that some
income sources are becoming more important.
Over time, there has been a large drop in the num-
ber of people generating income through farming.
In IRM-1, conducted shortly after the earthquakes,
68% said farming their own land was a major income
source.* 16 By IRM-2, conducted in February 2016,
this had declined to 51%. Since then, the proportion
reporting farming their own land as a main income
source has stayed fairly steady and is at 53% in IRM-
4 (Figure 3.1). There has been a similar decrease in
the number of people farming others’ land (from 6%
in IRM-1 to 3% in IRM-4). While livestock farming
dropped sharply as an income source in the first year
after the earthquakes, it has since almost recovered
with 18% reporting it as a major income source in the
latest survey.
In contrast, far more people are generating income
through their own business or daily wage work
than in the past and remittances have become more
important. Whereas 23% of people cited business
revenue as a major income source in June 2015, this
had increased to 36% by April 2017. Eight percent of
people said daily wage work was an important income
source in June 2015; this increased to 17% in the latest
survey. Those citing remittances as a major income
source increased from 10% in June 2015 to 15% in
April 2017.
To what extent have income sources improved
in the last three months?
Most people continue to see improvements in
their income sources but the proportion seeing
improvements in the past three months has declined
for most sources compared to IRM-3. Overall, 58%
of respondents said that at least one of their income
sources was affected by the earthquake. Of these
people, 83% said that at least one of their income
sources had improved in the last three months. The
proportion citing recent improvements is high for
every income source.
For most income sources, however, the proportion
saying they had seen recent improvements is lower
than was the case for IRM-3 (Figure 3.2). Seventy-
two percent of people who farm their own land whose
income was affected by the earthquakes, for example,
say they have seen recent improvements compared to
85% in IRM-3.
Daily wage work, business income and remittances are
the exceptions: for each, a much larger share of people
say this income source has improved in the last three
months compared to IRM-3. This may explain why
more people are now relying on these three income
sources than was the case before.
16 Respondents in IRM-4 could choose more than one option from
16 different income sources.
26


Aid and Recovery in Post-Earthquake Nepal
Figure 3.1: Income sources for people in affected areas (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Farming one's own land
Farming another's land
Daily wages
Own business
Remittance
Private company
Government service
Pension
Rent
Livestock farming
Jun 2015 (IRM-1)
Feb-Mar 2016 (IRM-2)
â–  Sep 2016 (IRM-3)
Apr 2017 (IRM-4)
Figure 3.2: Share of people within each income source whose income from that source
has improved (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2)
| Apr 2017 (IRM-4)
Sep 2016 (IRM-3)


Shelter
Changes in levels of income
Around one-third of people say their current income
is lower than before the earthquakes but a significant
proportion also say it has increased. Looking only at
those who were interviewed in the last three rounds of
the survey, 34% of people in IRM-4 report that their
current income is lower than their pre-earthquake
income (Figure 3.3).17 Twenty-seven percent report a
higher income than before the earthquakes while 38%
say that their income has not changed.
People in severely hit and crisis hit districts are more
likely to have seen a decline in income compared to
those in lower impact districts. On average, individuals
in severely hit or crisis hit districts are nearly 15 per-
centage points more likely than those in lesser affected
districts to report that their income has declined since
the earthquakes (Table 3.1).18
Declining income has been particularly widespread in
Sindhupalchowk and Gorkha districts. In each, over
half of the population report that their current income
is lower than their income before the earthquakes and
the rate is also high in Okhaldhunga (46%).
Figure 3.3: Current income (IRM-4)
compared to pre-earthquake income (IRM-2)
(IRM-4, IRM-2 household panel, unweighted)
Table 3.1: Current income (IRM-4) compared to pre-earthquake income (IRM-2) -
by district impact and district (IRM-4, IRM-2 household panel, unweighted)
No change Income decreased Income increased Don’t know
Severely hit 32% 38% 27% 3%
Dhading 38% 25% 37% 1%
Gorkha 27% 51% 20% 2%
Nuwakot 34% 42% 20% 4%
Ramechhap 37% 20% 39% 5%
Sindhupalchowk 28% 52% 18% 2%
Crisis hit 36% 35% 25% 5%
Bhaktapur 35% 22% 32% 11%
Kathmandu 32% 23% 38% 7%
Okhaldhunga 37% 46% 16% 0%
Hit with heavy losses 33% 22% 28% 17%
Lamjung 43% 20% 36% 1%
Solukhumbu 17% 25% 14% 44%
Hit 24% 21% 28% 26%
Syangja 24% 21% 28% 26%
All districts 38% 34% 27% 1%
17 Data on pre-earthquake income is taken from the IRM-2 survey. 18 This is based on unweighted estimates from the panel dataset.
28


Aid and Recovery in Post-Earthquake Nepal
Income recovery in more remote areas is lagging
behind that in other regions. A much larger share of
people in more remote areas say that their income
has declined since the earthquakes (Figure 3.4). Fif-
teen percentage points more people in more remote
districts say their income has declined compared to
people in less remote areas, while 14 percentage points
more people in less remote areas say their income has
improved compared to those in more remote areas.
Figure 3.4: Current income (IRM-4) compared to
pre-earthquake income (IRM-2) - by remoteness
Less remote Remote More remote
No change H Income increased
Income decreased H Don't know
Income recovery and housing damage.
Unsurprisingly, people who sustained greater damage
to their houses are more likely to struggle with
income recovery. As shown in Figure 3.5, those whose
house was completely destroyed are 14 percentage
points more likely to have seen their income decline
compared with those whose house was not damaged
and 11 points more than those whose house suffered
minor damage. People whose house saw minor or no
damage are more likely to have seen their income
increase than decrease.
Figure 3.5: Current income (IRM-4) compared
to pre-earthquake income (IRM-2) - by housing
Completely Major Minor Not
destroyed damage damage damaged
No change H Income increased
Income decreased Don't know
Income recovery and where people are
currently living. People living in shelters are more
likely to have a lower level of income than before the
earthquakes compared to those who are now living in
their own house (Table 3.2).
People who live in their own house are 5 percentage
points more likely to report improvements in income
since the earthquakes than are people living in shelters
on their own land. Those living in such shelters are
7 percentage points more likely to report that their
income has declined. Similarly, those living in shelters
on other people’s land or who are renting are also
more likely to have experienced a decline in income
compared to people living in their own homes.
Table 3.2: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - by where people are
living (IRM-4, IRM-2 household panel, unweighted)
No change Income declined Income improved Don’t know/ refused
Own house 31% 31% 29% 10%
Neighbor’s house* 35% 39% 19% 6%
Self-constructed shelter on own land 35% 38% 24% 3%
Self-constructed shelter on other people’s land 26% 54% 19% 1%
Self-constructed shelter on public land* 40% 50% 10% 0%
Community shelter* 0% 100% 0% 0%
Rent 57% 24% 14% 5%
★Less than 1 %
29


Shelter
Differences in income recovery across groups
Social structures and networks often play an important
role in the recovery of livelihoods after exogenous
shocks. Disaggregating the data on changes in income
by people’s identity and past income shows that those
who were poorer before the earthquake, or who come
from less privileged social groups, are much less likely
to have seen their income recover than others.
People’s caste appears to have played a small role in
the likelihood that their income has recovered. Recov-
ery among Janajatis is lower with them being slightly
more likely to report that their income has declined
since the earthquakes compared to high caste and low
caste people (Table 3.2). A higher share of high caste
people say their income has improved (28%), com-
pared to Janajatis (26%) and low caste people (24%).
People’s initial pre-earthquake income is a more power-
ful determinant of income recovery. While 58% of those
who had a low income before the earthquakes report
that their income has declined since then, 69% of those
who had a high income before the earthquakes say that
their income has improved in the past two years.
There is little difference in changes in income between
men and women.19 However, widows and those with a
disability are slightly more likely to report that their
income has declined. Widows are 2 percentage points
more likely and people with disabilities are 3 points
more likely than others to have seen their income
decline.
Table 3.3: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - by caste,
pre-earthquake income, widows and disability (IRM-4, IRM-2 household panel, unweighted)
No change Income declined Income increased Don’t know
High caste 34% 31% 28% 7%
Caste Janajati 32% 36% 26% 7%
Low caste 35% 32% 24% 8%
Pre-earthquake income Low 34% 58% 7% 1%
Medium 42% 15% 41% 2%
High 21% 7% 69% 2%
Widows
Widows 29% 36% 27% 8%
Disability No disability 33% 34% 27% 7%
Disability 31% 37% 22% 11%
Education level is also linked to changes in income
since the earthquakes. More educated individuals
are more likely to have seen their income improve,
while less educated individuals are more likely to
have experienced a decrease in income. As shown in
Table 3.4, compared to individuals with no education,
a person with the highest level of education is two
times more likely to report increased income. Those
with no education are almost three times more likely
to report a decline in income.
Table 3.4: Current income (IRM-4) compared to pre-earthquake income (IRM-2) - by education
(IRM-4, IRM-2 household panel, unweighted)
No change Income declined Income improved Don’t know/ refused
Illiterate 31% 38% 22% 9%
Literate 35% 37% 24% 4%
Primary level 36% 29% 27% 8%
Education level Lower secondary level 31% 29% 33% 6%
Secondary level 28% 30% 33% 10%
SLC Pass 34% 28% 33% 5%
+2/lntermediate pass 29% 27% 38% 7%
Bachelor pass 34% 22% 40% 4%
Master and above 23% 14% 45% 18%
30


Aid and Recovery in Post-Earthquake Nepal
3.2 Food
Current and future need for food
Far fewer people are saying that the provision of food
is a priority need for them than in previous IRM sur-
veys.19 20 Only 7% of the population in IRM-4 say that
food is one of their most important immediate needs,
down from 27% in IRM-1, and 7% that it is an impor-
tant need for next three months, down from 24% in
IRM-1. The drop in food demand has been steady since
IRM-1 (Figure 3.6).
Figure 3.6: Food as a top immediate need and three month need (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Immediate food need I Next 3 months food need
However, food continues to remain an acute need in severely hit districts are much more likely than others
some areas and for certain groups of people. Those in to say they need food (Figure 3.7).
Figure 3.7: Food as a top immediate need and three month need - by district impact (IRM-4, weighted)
Foor need immediate Food need next 3 months
19 Thirty-four percent of both men and women say their income 20 Food in the survey mainly refers to rice, wheat and maize, which
has declined. Twenty-seven percent of men say their income has are the main staple foods in Nepal.
increased compared to 26% of women.
31


Shelter
Table 3.5: Food as a top immediate need and
three month need - by district impact and district
(IRM-4, weighted)
Immediate food need 3 month food need
Severely hit 20% 24%
Dhading 4% 3%
Gorkha 38% 28%
Nuwakot 20% 45%
Ramechhap 9% 11%
Sindhupalchowk 26% 34%
Crisis hit 3% 1%
Bhaktapur 7% 3%
Kathmandu 1% 1%
Okhaldhunga 15% 7%
Hit with heavy losses 8% 8%
Lamjung 4% 5%
Solukhumbu 14% 14%
Hit 11% 6%
Syangja 11% 6%
All districts 7% 7%
Stated need for food is particularly high in Gorkha,
Sindhupalchowk and Nuwakot (Table 3.5). The
demand for food in districts that are more urban,
such as Kathmandu and Bhaktapur, is much lower
compared to other districts.
Food demand is also much higher in more remote areas
than in remote or less remote areas (Figure 3.8).21 Two
times as many people in more remote areas (27%) say
food is an important immediate need than do those
in remote areas, and 50 percent more people say it is
a three month need. Reported food need is around
seven times higher in more remote areas (both as the
most important immediate need and for the next three
months) compared to less remote areas.
Figure 3.8: Food as a top immediate need
and three month need - by remoteness
(IRM-4, weighted)
Foor need immediate I Food need next 3 months
Differences in food need across groups
Stated need for food is substantially higher for those
of low caste or who had a low pre-earthquake income.
Immediate food need is almost three times higher
among low caste people (13%) and nearly seven
times higher among the low income group (14%)
compared to high castes (4%) and those with a high
pre-earthquake income (2%) (Table 3.6).
Table 3.6: Food as a top immediate need and three month need - by caste and
pre-earthquake income (IRM-4, weighted)
Food need immediate Food need next 3 months
High caste 4% 4%
Caste Janajati 8% 7%
Low caste 13% 7%
Pre-earthquake income Low 14% 13%
Medium 4% 4%
High 2% 1%
21 Since most places in crisis hit districts are less remote (78%), this
is the main factor driving the low food demand in these districts.
32


Aid and Recovery in Post-Earthquake Nepal
There is not much variation in stated food need by
gender.22 However, widows are 5 percentage points
more likely to say they have immediate food needs
and 2 percentage points more likely to report food
need for the next three months (Table 3.7). Similarly,
people with a disability are twice as likely as others
to say they have an immediate need for food and 1.5
times more likely than others to mention food need
for the next three months.
Table 3.7: Food as a top immediate need
and three month need - by widows and disability
(IRM-4, weighted)
Food need immediate Food need next 3 months
Widows 11% 8%
No disability 6% 6%
Disability 13% 9%
Have food prices increased?
When asked if food prices have changed since the last
monsoon, 66% say the price of at least one type of food
has become much higher, 80% say the price of at least
one type has become slightly higher and 28% mention
that the price of at least one type of food has remained
the same (Figure 3.9), while nearly 3% mention food
that at least one type of food is now less costly (not
shown) and around 50% have either not bought a
given type of food or do not know (not shown).23 Food
prices appear to have increased more drastically in
higher impact districts, as an average of 69% people
in the top two impact categories say that food prices
have become much higher compared to 47% in the
lower two impact categories. However, the lower two
impact categories have also experienced a rise in food
prices with a higher share of people in the bottom two
impact categories mentioning that food prices have
slightly increased. This suggests that price increases
have been common across all districts but are more
severe in higher impact districts.
Figure 3.9: Increase in food prices - by district impact (IRM-4, weighted)
Much higher Slightly higher H Same
More remote districts have the highest proportion of
respondents (68%) who report much higher increase
in food prices. But the remote category has the highest
share of people who say food prices have increased
slightly (Figure 3.10).
22 This is not surprising given people are likely reporting on food
need for their households.
23 Respondents were asked for changes in food prices for various
items such as rice, wheat, lentils, cooking oils, vegetable, meat and
farm products. Mentioned percentages represent changes in food
prices as an aggregate measure from these multiple set responses.
Therefore, responses add to more than 100%.
33


Shelter
Figure 3.10: Increase in food prices - by remoteness (IRM-4, weighted)
Much higher H Slightly higher H Same
Changes in food consumption
There do not appear to be widespread decreases in
food consumption. Respondents in each survey were
asked to compare their current food consumption
level with that a year before. As with previous rounds
of the survey, most people say their year-on-year
consumption has remained the same, with around
one-third saying it has increased and 6% reporting a
decrease (Figure 3.11). Thirty-four percent now say
their consumption has increased over the past year
compared to 31% in February 2016. This matches with
an improvement in food security reported in the latest
Nepal Food Security Monitoring system (NeKSAP),
which covers November 2016-March 2017.24
Figure 3.11: Food consumption compared to last year (IRM-2, IRM-4, weighted)
Apr 2017 (IRM-4) H Feb-Mar 2016 (IRM-2)
24 The NeKSAP detailed report (Issue 50) has yet to be published,
but the preliminary map of the report can be found at http://
neksap.org.np/home [accessed June 20, 2017].
34


Aid and Recovery in Post-Earthquake Nepal
When asked about changes in food consumption say their consumption has increased but are also more
since the end of the last monsoon (around August- likely to say it has decreased (Figure 3.12).
September 2016), respondents are now more likely to
Figure 3.12: Changes in food consumption in the past eight months (IRM-2, IRM-3, IRM4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Where is food consumption decreasing?
As shown in Table 3.8, there is substantial variation in (
changes to food consumption across districts. Those p
in severely hit districts are substantially more likely c
to report decreases in consumption. The four districts r
where more than 10% of people report decreases in food d
consumption are Okhaldhunga (13%), Sindhupalchowk
(12%), Nuwakot (13%) and Dhading (12%). Sixty-eight
percent of people in Nuwakot report an increase in food
consumption. There are no substantial differences in
reported changes to food consumption across areas of
different degrees of remoteness.
Table 3.8: Changes in food consumption in the past eight months - by district impact,
district and remoteness (IRM4, weighted)
Increased a lot Increased slightly Same as before Decreased slightly Decreased a lot Don’t know
Severely hit 4% 33% 54% 8% 1% 0%
Dhading 0% 15% 73% 12% 0% 0%
Gorkha 11% 28% 59% 3% 0% 0%
Nuwakot 4% 64% 17% 8% 5% 2%
Ramechhap 1% 27% 71% 2% 0% 0%
Sindhupalchowk 2% 36% 50% 12% 0% 0%
Crisis hit 3% 21% 70% 5% 1% 0%
Bhaktapur 1% 29% 64% 5% 1% 1%
Kathmandu 3% 20% 72% 5% 1% 0%
Okhaldhunga 2% 19% 65% 12% 1% 0%
Hit with heavy losses 3% 30% 63% 3% 0% 0%
Solukhumbu 9% 18% 68% 5% 0% 0%
Lamjung 1% 37% 61% 1% 0% 0%
Hit 4% 36% 54% 5% 1% 0%
Syangja 4% 36% 54% 5% 1% 0%
Less remote 3% 24% 66% 6% 1% 0%
Remote 3% 28% 62% 6% 1% 0%
More remote 7% 30% 56% 8% 0% 0%
35


Shelter
Whose food consumption is decreasing?
A higher share of low caste and Janajati people report h
decreasing food consumption (Table 3.9). However, ii
levels of pre-earthquake income are more important, d
Those in the low income group are 3 percentage points a
more likely than those in the medium income group n
and 7 points more likely than those in the high income d
group to report a decrease in food consumption in the
last eight months. There is also a noticeable difference
in reported decreases in food consumption when
disaggregating by gender too. Women (9%) are twice
as likely to report a decrease in consumption as are
men (4%). The difference by education, widows and
disability is not clear or large.
Table 3.9: Changes in food consumption in the past eight months - by caste,
pre-earthquake income and gender (IRM4, weighted)
Increased a lot Increased slightly Same as before Decreased slightly Decreased a lot Don’t know/ refused
High caste 3% 26% 65% 4% 1% 0%
Caste Janajati 3% 27% 62% 7% 0% 1%
Low caste 4% 26% 62% 7% 1% 0%
Pre-earthquake income Low 3% 29% 56% 9% 1% 1%
Medium 4% 24% 65% 6% 1% 0%
High 2% 26% 69% 3% 0% 0%
Gender Female 3% 27% 61% 8% 1% 0%
Male 3% 26% 66% 4% 0% 0%
Food consumption among groups in Okhaldhunga,
Sindhupalchowk, Nuwakot and Dhading
These four districts have higher shares of people
who reported a decrease in food consumption.
Table 3.10 provides analysis by different groups for
these districts only.
Findings are similar to those in other districts. While
all caste and income group categories in these districts
are more vulnerable compared to other districts,
decreases in food consumption is more pronounced
among those of low caste. Individuals in the low
caste group are 3 percentage points more likely than
Janajatis and 4 points more likely than the high caste
group to experience a decrease in food consumption.
Table 3.10: Changes in food consumption in the past eight months in Okhaldhunga, Sindhupalchowk,
Nuwakot and Dhading - by caste and pre-earthquake income (IRM4, weighted)
Increased a lot Increased slightly Same as before Decreased slightly Decreased a lot Don’t know/ refused
High caste 1% 29% 58% 10% 2% 1%
Caste Janajati 2% 37% 47% 12% 1% 1%
Low caste 4% 25% 56% 12% 4% 0%
Pre-earthquake income Low 2% 33% 50% 12% 2% 1%
Medium 2% 36% 54% 8% 0% 0%
High 1% 35% 50% 14% 0% 0%
36


Aid and Recovery in Post-Earthquake Nepal
3.3 Public services
As highlighted in earlier IRM reports, access to public
services has improved since the immediate aftermath
of the disaster. In particular, access to drinking water
improved markedly between June 2015 and March
2016. However, since then, there have not been sig-
nificant changes in the proportion of people reporting
they have access to most services. The one exception
is access to drinking water. Nine percentage points
fewer people say they have access to drinking water in
April 2017 compared to September 2016 (Figure 3.13).
Figure 3.13: Share saying they have services provided by the VDC/municipality
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) | Sep 2016 (IRM-3)
Feb-Mar 2016 (IRM-2) H Apr 2017 (IRM-4)
Regions that have less access to drinking water
Access to clean drinking water has declined in severely
hit and crisis hit districts. As shown in Table 3.11, while
only 46% of people in severely hit districts reported
having access to clean drinking water immediately
after the earthquake (IRM-1), the figure increased to
75% in IRM-2 and 85% in IRM-3. However, only 76%
of the severely hit population said they had access to
clean drinking water in IRM-4.
The result is driven mainly by two districts, Gorkha
and Nuwakot, where reported water accessibility in
IRM-4 has lowered by 19 percentage points and 29
points, respectively, since IRM-3. In crisis hit districts,
there has been a decline of 12 points from IRM-3 to
IRM-4. Among crisis hit districts, the share of people
with access to clean water declined by 7 percentage
points in Bhaktapur and by 13 points in Kathmandu.
There is no change in water accessibility, however, in
the other two impact categories. November to June
is a relatively dry period in Nepal and it is likely that
the impact of the disaster on infrastructure combined
with the dry weather has exacerbated problems with
accessing clean drinking water in IRM-4.
Table 3.11: Access to clean drinking water - by district impact and district
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Severely hit 46% 75% 85% 76%
Dhading 36% 71% 77% 82%
Gorkha 32% 91% 87% 68%
Nuwakot 74% 71% 95% 66%
37


Shelter
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Ramechhap 59% 76% 80% 86%
Sindhupalchowk 36% 68% 85% 81%
Crisis hit 83% 94% 92% 80%
Bhaktapur 73% 92% 89% 82%
Kathmandu 86% 95% 93% 80%
Okhaldhunga 66% 86% 76% 81%
Hit with heavy losses 82% 95% 99% 100%
Solukhumbu 94% 91% 100% 100%
Lamjung 75% 97% 99% 100%
Hit 51% 90% 97% 97%
Syangja 51% 90% 97% 97%
All districts 69% 88% 90% 81%
Satisfaction with public services
Satisfaction rates with public services have declined
in IRM-4 with the exception of electricity where more
people are satisfied than in the past (Table 3.12).
Highest levels of dissatisfaction are with drinking
water (23%) and roads (15%). Levels of dissatisfaction
with water is high in both IRM-2 and IRM-4, both of
which cover the dry season.
Table 3.12: Satisfaction with public services (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Satisfied 89% 60% 63% 91%
Electricity Neither satisfied nor dissatisfied 5% 5% 18% 5%
Dissatisfied 6% 35% 19% 4%
Satisfied 85% 61% 67% 62%
Drinking water Neither satisfied nor dissatisfied 7% 6% 17% 15%
Dissatisfied 8% 33% 16% 23%
Satisfied 93% 81% 80% 67%
Medical facilities Neither satisfied nor dissatisfied 4% 5% 13% 24%
Dissatisfied 3% 14% 7% 9%
Satisfied 93% 85% 90% 77%
Schools Neither satisfied nor dissatisfied 3% 4% 7% 20%
Dissatisfied 4% 11% 3% 3%
Satisfied 90% 80% 80% 64%
Motorable roads Neither satisfied nor dissatisfied 4% 5% 10% 21%
Dissatisfied 6% 15% 10% 15%
38


Aid and Recovery in Post-Earthquake Nepal
Where are people more dissatisfied with
public services in April 2017?
Rates of dissatisfaction differ greatly between districts, o
Dissatisfaction with electricity in highest in the n
severely hit districts but is much lower in Dhading e:
and Nuwakot than elsewhere (Table 3.13). People w
are most dissatisfied with electricity in Okhaldhunga. w
Dissatisfaction with drinking water is high everywhere, B
regardless of earthquake impact, with the exceptions
of Nuwakot, Solukhumbu and Lamjung. Similarly,
many people are dissatisfied with medical facilities
except in Nuwakot and Kathmandu. Dissatisfaction
with schools is highest in Solukhumbu. Dissatisfaction
with roads is much higher in Gorkha, Solukhumbu and
Bhaktapur than elsewhere.
Table 3.13: Dissatisfaction with public services - by district impact and district (IRM-4, weighted)
Electricity Drinking water Medical facilities Schools Motorable roads
Severely hit 8% 22% 16% 5% 15%
Dhading 4% 23% 12% 2% 10%
Gorkha 17% 24% 26% 7% 35%
Nuwakot 0% 3% 1% 4% 6%
Ramechhap 4% 28% 24% 5% 15%
Sindhupalchowk 12% 29% 15% 7% 9%
Crisis hit 2% 27% 3% 1% 16%
Bhaktapur 0% 27% 15% 6% 23%
Kathmandu 1% 29% 1% 1% 16%
Okhaldhunga 24% 12% 9% 4% 4%
Hit with heavy losses 3% 2% 15% 9% 13%
Solukhumbu 4% 1% 22% 21% 28%
Lamjung 2% 3% 11% 2% 8%
Hit 2% 23% 20% 1% 10%
Syangja 2% 23% 20% 1% 10%
All districts 4% 23% 9% 3% 15%
For most services, people in more remote areas are The one exception is drinking water where people in
more likely to be dissatisfied than others (Figure 3.14). less remote areas are far more likely to be dissatisfied.
Figure 3.14: Dissatisfaction with public services - by remoteness (weighted, IRM-4)
Less remote Remote H More remote
39


Shelter
Which groups are dissatisfied with public services?
As shown in Table 3.14, low caste people are generally
more likely to be dissatisfied with most public services,
although not by a large margin. A notable exception
is drinking water, with 31% of low caste people ex-
pressing dissatisfaction compared to 25% of high caste
people and 22% of Janajatis. This reflects ongoing
problems many low caste people have in accessing
public water supplies.
Those with a low pre-earthquake income are more
likely to express dissatisfaction with electricity,
medical facilities and schools, but less likely to be
dissatisfied with drinking water and roads. Those with
a high income are most likely to be dissatisfied with
the quality of roads.
Table 3.14: Dissatisfaction with public services - by caste and pre-earthquake income (IRM-4, weighted)
Electricity Drinking water Medical facilities Schools Roads
High caste 3% 25% 9% 3% 16%
Caste Janajati 4% 22% 9% 3% 15%
Low caste 5% 31% 12% 4% 14%
Low 7% 20% 14% 5% 12%
Pre-earthquake income Medium 3% 26% 6% 2% 13%
High 2% 24% 6% 2% 20%
Women are more likely than men to be dissatisfied
with drinking water but less likely to dissatisfied with
roads (Table 3.15). Widows are particularly likely to
be dissatisfied with drinking water. The disabled are
more likely to be dissatisfied with most public services
but are particularly likely to be dissatisfied with health
facilities.
Table 3.15: Dissatisfaction with public services - by gender, widows and disability
(Apr 2017, IRM-4, weighted)
Electricity Drinking water Medical facilities Schools Roads
Female 4% 25% 9% 3% 12%
Male 4% 21% 9% 3% 18%
Widows 4% 31% 12% 5% 13%
No disability 4% 23% 8% 3% 15%
Disabled 6% 23% 23% 6% 19%
40


Photo: Chiran Manandhar
This chapter examines the coping strategies used
by people in affected areas to deal with the impacts
of the earthquake and to recover. These include
borrowing, sale of assets, remittances and migration.
Previous rounds of IRM found that borrowing is the
most common strategy. As such, the chapter focuses
primarily on borrowing, looking at who is taking loans
from whom and why, how this has changed over time
and levels of debt.
Key Findings
Borrowing
• Two years after the disaster, borrowing continues
to increase in affected districts. Borrowing has
increased most sharply in more affected districts.
Fifty-five percent of people have borrowed in the
last eight months in the severely hit districts,
compared to 24% in the early months after the
earthquake.
• A larger proportion of people in more remote
areas are borrowing than elsewhere. Compared
to 39% people in less remote areas, 47% of people
in remote areas and 55% in more remote areas
borrowed in IRM-4.
• As in previous surveys, those who had a low
income before the earthquake and individuals of
low caste are more likely to borrow than others.
Borrowing in IRM-4 has also increased among
people with disabilities.
• People who sustained greater damage to their
house are also more likely to borrow, and they are
more likely than others to borrow for rebuilding.
• People in more remote areas are borrowing from
informal sources such as moneylenders, friends,
relatives and neighbors. These informal sources
typically charge higher interest rates. In contrast,
people in less remote areas are borrowing more
from formal sources: banks, savings and credit
organizations, cooperatives and other financial
institutions.
• A higher share of people in higher impact districts
and more remote areas are regular borrowers.
They are also more likely to borrow in future.
Those in more remote regions, and in more
affected areas, are at greater risk of falling into
debt traps.
Asset sales
• Sales of assets have increased and are highest in
more affected districts. While only 4% of people
said they sold assets in IRM-2, and 3% in IRM-3,
6% now report having sold assets in the last eight
months. Sale of assets remains most common in
the severely hit districts.
41


Coping Strategies
Photo: Nayan Pokharel
• The majority of people who sold assets in IRM-4
sold land (43% of those who sold assets) or livestock
(40%). People have also sold gold (9%) or their
house (5%) to cope with the earthquakes’ impacts.
• Data confirm the earlier finding that borrowing
frequency is associated with the likelihood of
asset sales. Those who have borrowed repeatedly
since February 2016 (IRM-2) are more than twice
as likely as those who have not borrowed in any
of the last three waves of the survey to sell assets
to cope with earthquake impacts.
• Compared to those living in their own houses
(6% of whom have sold assets), a slightly higher
proportion of people living in shelters on their
own land (9%) or on other’s land (8%) sold assets
in IRM-4.
Remittances
• Remittances are becoming more important as a
source of income. Fifteeen percent of people in
affected areas say remittances are one of their
main income sources in IRM-4, compared to 9%
in IRM-1. However, remittances still tend to be
more important in less affected districts and for
those with a high income.
• The level of housing damage and current housing
conditions do not correlate with the likelihood of
them receiving remittances.
Migration
• Most people say levels of out-migration from
their communities have stayed the same as before
the earthquakes. However, more people say that
migration has increased than decreased. Overall,
65% of people say migration levels have remained
the same, 20% say they have increased, and 4%
say levels have decreased.
• There is no clear pattern in reported migration
by the level of earthquake impact but increases
in reported out-migration are greater in less
remote areas.
• Plans for migration in the next year suggest the
earthquakes have an influence as a majority of
those who plan to do so (61%) are from severely
hit districts.
42


Aid and Recovery in Post-Earthquake Nepal
4.1 Borrowing
Changes in borrowing over time
Two years after the disaster, borrowing continues to
increase in earthquake-affected districts. In the im-
mediate aftermath of the earthquakes, 14% of people
borrowed money. Thirty-two percent took loans in
IRM-2 and the rate stayed the same in IRM-3. The
amount of people borrowing has grown further in
the past eight months with 44% having taken loans
in this period.
Borrowing has generally increased most sharply in
more affected districts (Table 4.1).25 26 Fifty-five percent
of people have borrowed in the last eight months
in the severely hit districts, compared to 24% in
the early months after the earthquake. Borrowing
is now particularly high in the severely hit districts
of Dhading and Ramechhap as well as the crisis hit
district of Okhaldhunga, where 72% of people have
taken loans in IRM-4. Borrowing is least common in
Lamjung district.
Table 4.1: Share of people who have borrowed - by district impact and district
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)26
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Severely hit 24% 49% 43% 55%
Dhading 25% 52% 48% 64%
Gorkha 17% 45% 36% 52%
Nuwakot 14% 43% 34% 54%
Ramechhap 40% 63% 59% 55%
Sindhupalchowk 30% 46% 42% 49%
Crisis hit 11% 22% 25% 39%
Bhaktapur 11% 22% 14% 40%
Kathmandu 9% 19% 23% 36%
Okhaldhunga 30% 66% 66% 72%
Hit with heavy losses 10% 24% 24% 21%
Lamjung 7% 21% 23% 18%
Solukhumbu 15% 29% 26% 27%
Hit 4% 43% 45% 51%
Syangja 4% 43% 45% 51%
All districts 14% 32% 32% 44%
25 It has also increased in Syangja, the least affected district. It is
unclear why this is the case.
26 The time period covered in this survey question differs slightly for
each survey round. Respondents were asked if they had borrowed
since the earthquake in IRM-1, since the beginning of last
monsoon in IRM-2 (June 2015-February 2016), since the end of
winter season in IRM-3 (March 2016-September 2016), and since
the end of last monsoon in IRM-4 (September 2016-April 2017).
43


Coping Strategies
A larger proportion of people in more remote areas 47% of people in remote areas and 55% in more remote
are borrowing than elsewhere. As shown in Figure 4.1, areas borrowed in IRM-4.
Figure 4.1: Share of people who have borrowed - by remoteness (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) â–  Sep 2016 (IRM-3)
Feb-Mar 2016 (IRM-2) H Apr 2017 (IRM-4)
Who is borrowing?
Income. As in previous surveys, those who had a low
income before the earthquake are more likely to have
borrowed since the earthquakes than others. There has
been a steady increase in the share of low income people
who are borrowing over time - from 35% in IRM-2 to
40% in IRM-3 to 52% in IRM-4 (Figure 4.2). Borrowing
for those who had a medium or high income has also
increased but still lags behind borrowing by the poor.
Figure 4.2: Share of people who have borrowed - by pre-earthquake income
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Caste. Compared to previous surveys, borrowing in
IRM-4 has increased for all caste groups (Figure 4.3).
But as in the past, a higher share of people of low caste
are taking loans. The proportion of high caste and low
caste people who borrowed in IRM-4 has increased by 16
percentage points since IRM-3 while the proportion of
Janajatis borrowing has increased by 9 points. However,
while 49% of those in the high caste group and 39% of
Janajatis are borrowing in IRM-4, nearly 62% in the
low caste say that they borrowed since the last survey.
44


Aid and Recovery in Post-Earthquake Nepal
Figure 4.3: Share of people who have borrowed - by caste (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Disability. There has been a steady increase in
borrowing by people with disabilities. As shown in
Figure 4.4, borrowing among people with disabilities
was slightly less frequent than borrowing by others in
IRM-2 and IRM-3. But the proportion of the disabled
who are taking loans surpassed others in IRM-4 by 6
percentage points.
Figure 4.4: Share of people who have borrowed -
by disability (IRM-2, IRM-3, IRM-4, weighted)
60%
50%
40%
30%
20%-------------1-----------1---------
Feb-Mar 2016 Sep 2016 Apr 2017
(IRM-2) (IRM-3) (IRM-4)
Table 4.2: Share of people who have borrowed - by
occupation (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Labor 32% 45% 44%
Unemployed 8% 41% 41%
Agriculture 43% 39% 50%
Service 36% 31% 42%
Housewife/ house-maker 27% 25% 32%
Industry/business 22% 23% 39%
Retired 11% 19% 20%
Student 26% 15% 61%
Housing damage. As in earlier surveys, people
whose houses were fully damaged are the most likely to
borrow (Figure 4.5). There has been a sharp increase
in the share of people whose house was destroyed who
are borrowing with 52% borrowing in the last eight
months, a 13 percentage point increase since IRM-3.
Borrowing has also risen sharply for those whose
house experienced minor damage.
No disability Disability
Occupation. Borrowing has increased the most
for those working in agriculture and those who are
students (Table 4.2). People in agriculture are 11
percentage points more likely to borrow in IRM-4
compared to IRM-3. Students are 46 percentage points
more likely to borrow than in IRM-4 than in IRM-3.27
27 Students make up less than 2% of the population in the last three
surveys.
45


Coping Strategies
Figure 4.5: Share of people who have borrowed - by housing damage (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Where people live. Those who still live in temporary borrowing has increased for those in all types of
shelters are the most likely to borrow although accommodation (Figure 4.6).
Figure 4.6: Share of people who have borrowed - by where people live (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2)
Sep 2016 (IRM-3)
| Apr 2017 (IRM-4)
Reasons for borrowing
The most common reason for borrowing is to support
livelihoods (60% of those who borrowed in IRM-4,an
increase from 58% in IRM-3). The next most common
reason is for people to rebuild their house (Figure 4.7).
There has been a sharp increase in borrowing for
housing reconstruction since IRM-3 but the share
of borrowers who took loans for this purpose is still
lower than in March 2016. The decline in the share of
people borrowing for food, for their business, or for
temporary shelter continues.
46


Aid and Recovery in Post-Earthquake Nepal
Figure 4.7: Reasons for borrowing, share of those borrowing (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
★Figure does not include responses that are 1 % or less
Figure 4.8: Share of those taking loans who are
borrowing for rebuilding - by housing damage
(IRM-4, weighted)
Borrowing for reconstruction. Borrowing to
rebuild houses is the second most cited reason for
borrowing in IRM-4 after livelihoods and it associates
well with earthquake impact levels (Table 4.3).
Compared to only 8% of borrowers in the hit district,
33% in severely hit districts, 23% in crisis hit ones
and 33% in hit with heavy losses districts say that
they borrowed money for rebuilding. Borrowing for
livelihoods accounts for a larger share of borrowers
in the crisis hit (67%) and hit districts (66%) than
in the severely hit (52%) and hit with heavy losses
(39%) districts. As in IRM-3, borrowing for food is
relatively high in both the severely hit (22%) and hit
districts (41%).
People who have suffered a higher level of damage to
their house are also more likely to borrow for rebuild-
ing when they take loans (Figure 4.8). Those whose
house was completely destroyed are 6 percentage
points more likely than those with major damage to
borrow for rebuilding. Unsurprisingly, they are near-
ly three times more likely to borrow for rebuilding
compared to people with minor or no damage to their
houses.
Table 4.3: Reasons for borrowing, share of those borrowing - by district impact (IRM-4, weighted)
Livelihoods Rebuild â– u 0 0 LL Temporary shelter Farming Improve shelter Education Treatment Business Migrate abroad Miscellaneous
Severely hit 52% 33% 22% 14% 11% 9% 4% 3% 3% 3% 4%
Crisis hit 67% 23% 9% 7% 2% 3% 1% 1% 3% 0% 2%
Hit with heavy losses 39% 33% 8% 16% 4% 7% 4% 1% 4% 3% 7%
Hit 66% 8% 41% 4% 11% 7% 7% 5% 5% 3% 3%
47


Coping Strategies
How much are people borrowing?
The average amount people borrowed has also
increased over time. As shown in Table 4.4, IRM-4
borrowers on average took loans of NPR 363,193, the
highest level since the earthquake and a threefold
increase since IRM-1. Borrowing in IRM-4 continues
to be the highest in the crisis hit districts, driven
mainly by the two urban districts of Kathmandu
and Bhaktapur. Among the severely hit districts, the
biggest increase in sums borrowed since IRM-3 has
been in Gorkha. Ramechhap and Lamjung are the two
districts where there has been a decline in the average
amount borrowed.
Table 4.4: Average borrowing in NPR - by district impact and district
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Severely hit 45,289 262,343 155,094 226,831
Dhading 54,719 645,171 172,533 234,771
Gorkha 53,910 149,389 152,641 255,675
Nuwakot 38,668 153,974 176,446 240,065
Ramechhap 44,811 118,267 121,906 199,719
Sindhupalchowk 34,859 111,245 150,104 192,695
Crisis hit 185,747 408,363 300,829 500,608
Bhaktapur 66,671 213,744 573,812 572,795
Kathmandu 243,843 531,259 324,193 543,756
Okhaldhunga 49,740 97,622 110,859 139,190
Hit with heavy losses 99,799 186,422 216,281 235,990
Lamjung 62,071 228,662 305,088 276,843
Solukhumbu 130,514 131,100 75,000 188,542
Hit 34,375 167,021 194,430 281,581
Syangja 34,375 167,021 194,430 281,581
All districts 103,057 303,130 213,451 363,193
The average loan size in less remote regions appears
to have spiked in IRM-2 but it continues to be the
higher than in other areas (Figure 4.9). The greatest
increase in loan size since IRM-3 is in remote areas,
where the average loan size in IRM-4 nearly doubled.
For more remote districts, loan sizes in IRM-4 have
increased compared to IRM-3, but are still lower than
the average loan size in IRM-2.
Figure 4.9: Average borrowing in NPR - by remoteness (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) | Sep 2016 (IRM-3)
Feb-Mar 2016 (IRM-2) H Apr 2017 (IRM-4)
48


Aid and Recovery in Post-Earthquake Nepal
Photo: Nayan Pokharel
Who are people borrowing from?
Cooperatives continue to be the most common source
of borrowing (27% of borrowers took loans from
cooperatives) - Table 4.5. Other common borrowing
sources in IRM-4 are relatives (19%), neighbors (17%)
and savings and credit groups (17%). Borrowing
from relatives, which was the most common source
immediately after the earthquake in IRM-1 (31%),
became less common in IRM-2 and IRM-3, but has
increased by 6 points since IRM-3. Borrowing from
banks has stayed constant at 13% since IRM-2.
Table 4.5: Sources of borrowing among those who borrowed
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Moneylender 13% 10% 12% 11%
Friend 12% 9% 7% 12%
Relative 31% 24% 13% 19%
Neighbor 18% 17% 19% 17%
Other individuals 1% 2% 1% 1%
Bank 2% 13% 13% 13%
Savings and credit group 17% 18% 20% 17%
Cooperatives 7% 15% 23% 27%
Other financial institutions 1% 5% 2% 2%
49


Coping Strategies
Formal and informal sources of borrowing.
The borrowing sources listed above can be divided into
informal sources (moneylenders, friends, relatives,
neighbors and other individuals) and formal sources
(banks, savings and credit groups, cooperatives and
other financial institutions). People in more remote
regions in IRM-4 borrowed largely from informal
sources (Figure 4.10). Those who borrow in less re-
mote regions are 16 percentage points more likely to
borrow from formal sources than those in remote areas
and 43% more likely than those in more remote areas.
In contrast, informal sources become more important
as remoteness increases. In more remote areas, bor-
rowers are 44 percentage points more likely to borrow
from informal sources than in less remote areas.
The average amount borrowed from formal sources
is highest in IRM-4, and there has also been a mod-
est increase in the average amount borrowed from
informal sources. The average amount borrowed
from banks increased from NPR 488,050 in IRM-3
to NPR 748,105 in IRM-4. Similarly, the average
amount borrowed in IRM-4 from cooperatives and
other financial institutions has doubled since IRM-3
(Table 4.6). Borrowing from friends and savings and
credit groups has declined since IRM-3. Worryingly,
average sums borrowed from moneylenders have
substantially increased.
Figure 4.10: Sources of borrowing among those
who borrowed - by remoteness (IRM-4, weighted)
Informal sources Formal sources
Table 4.6: Average borrowing in NPR - by sources (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Moneylender 66,009 763,730 107,966 183,618
Friend 55,080 99,064 462,343 210,138
Relative 156,562 102,836 208,144 217,525
Neighbor 123,576 103,889 103,631 148,955
Other individual from ward 24,534 97,546 154,018 165,779
Bank 87,196 887,654 488,050 748,105
Savings and credit group 53,888 109,503 98,616 92,985
Co-operatives 65,396 161,435 212,858 485,275
Other financial institution 11,522 130,528 48,458 119,346
Government loan scheme 12,696
What are monthly interest rates?
Average monthly interest rates have remained largely
steady since the earthquake. Interest rates charged by
informal sources, such as moneylenders, friends, rela-
tives and other individuals, have been, and continue to
be, higher than those charged by formal financial insti-
tutions (Figure 4.11). In IRM-4, interest rates charged
by banks, savings and credit cooperatives and other
financial institutions are 1.7-1.8 percent. For other infor-
mal sources, monthly interest rates range from 1.9-2.3
percent. People in more remote regions are depending
on informal sources despite the higher interest rates.
Earthquake impact and remoteness. More
than half of the people who took loans in IRM-4 say
that monthly interest rates were 1.5-2% and around
19% say that interest was more than 2%. As shown in
Table 4.7, higher interest rates are prevalent in higher
impact districts, although the less affected Lamjung
and Syangja districts are exceptions. Ramechhap
(32%) and Sindhupalchowk (31%) have the highest
shares of people who say they are charged more than
2% interest per month.
50


Aid and Recovery in Post-Earthquake Nepal
People in less remote regions are nearly three times
more likely than more remote regions to have monthly
interest rates of less than 1%. In contrast, people in
more remote regions are twice as likely as those in
less remote areas to be charged interest rates of more
than 2%.
Figure 4.11: Changes in interest rates from different sources
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1)
Feb-Mar 2016 (IRM-2)
â–  Sep 2016 (IRM-3)
Apr 2017 (IRM-4)
Table 4.7: Mean reported interest rates - by district impact,
district and remoteness (IRM-4, weighted)
Less than 1 % Between 1%-1.5% Between 1.5%-2% Above 2% Refused
Severely hit 15% 6% 56% 23% 1%
Dhading 13% 6% 69% 12% 1%
Gorkha 14% 8% 53% 23% 2%
Nuwakot 12% 2% 62% 23% 1%
Ramechhap 14% 4% 50% 32% 1%
Sindhupalchowk 24% 8% 36% 31% 1%
Crisis hit 30% 4% 48% 15% 3%
Bhaktapur 76% 1% 21% 1% 1%
Kathmandu 26% 4% 50% 17% 4%
Okhaldhunga 6% 5% 68% 21% 0%
Hit with heavy losses 24% 1% 56% 16% 2%
Lamjung 16% 2% 54% 27% 2%
Solukhumbu 34% 1% 58% 4% 2%
Hit 11% 3% 62% 22% 2%
Syangja 11% 3% 62% 22% 2%
All districts 22% 4% 52% 19% 2%
Less remote 34% 5% 43% 14% 4%
Remote 13% 4% 61% 20% 1%
More remote 13% 5% 48% 33% 1%
51


Coping Strategies
How have levels of debt changed?
As shown in Table 4.4 above, average debt size has
increased from NPR 103,057 in IRM-1 to NPR 363,193
in IRM-4. Debt loads are also increasing for a sub-
stantial share of the population. According to IRM-4,
overall debt has increased for 47% of the people who
took loans since the last monsoon (September 2016).
As shown in Figure 4.12, overall debt has decreased
for just 11% of the population who borrowed in the
past eight months.
Levels of debt correlate strongly with the degree of
earthquake impact districts experienced. Among those
who have taken a loan, 53% in severely hit districts
say that their overall debt has increased since IRM-3,
compared to 46% in crisis hit, 28% in hit with heavy
losses and 37% in hit districts (Table 4.8). The two
districts where debt increase is the highest among
those who have taken loan are Okhaldhunga (62%)
and Dhading (70%). Debt appears to be increasing
for more people in more remote areas (53%) than in
remote (51%) or less remote areas (41%).
Figure 4.12: Changes in debt (IRM-4, weighted)
Table 4.8: Changes in debt - by district impact, district and remoteness (IRM-4, weighted)
Increased No change Decreased Don’t know/refused
Severely hit 53% 33% 13% 1%
Dhading 70% 14% 15% 1%
Gorkha 57% 35% 7% 1%
Nuwakot 32% 54% 14% 0%
Ramechhap 58% 33% 8% 1%
Sindhupalchowk 42% 37% 21% 0%
Crisis hit 46% 43% 10% 1%
Bhaktapur 32% 38% 28% 2%
Kathmandu 46% 45% 8% 1%
Okhaldhunga 62% 28% 10% 0%
Hit with heavy losses 28% 56% 8% 9%
Solukhumbu 22% 73% 4% 1%
Lamjung 36% 31% 14% 20%
Hit 37% 48% 13% 2%
Syangja 37% 48% 13% 2%
Less remote 41% 46% 12% 1%
Remote 51% 35% 12% 2%
More remote 53% 41% 5% 1%
Debt increase is most common for people whose hous-
es were completely destroyed. As shown in Figure 4.13,
debt is increasing for 52% of those whose house was
completely destroyed, while it is for 40% of those
whose house saw major damage and for 45% of people
with minor damage to their houses. It is noteworthy
that 33% of those without any house damage have
increasing overall debt, which suggests that rising
debt is a wider problem and not solely related to the
earthquakes. This could be seen in the earlier analysis
which found that borrowing was prevalent not only
among people with housing damage, but also among
others. However, the fact that those who were the
most affected are the most likely to borrow and to
have rising debt suggests that the earthquake has led
to more borrowing and debt.
52


Aid and Recovery in Post-Earthquake Nepal
Figure 4.13: Increase in debt - by housing damage
(IRM-4, weighted)
Caste and income. Rising debt is most common
for those of a high caste. Half of this group say that
their overall debt is increasing, compared to 46% of
Janajatis and 37% of low caste people who have taken
loans since IRM-3 (Table 4.9). When disaggregating
by pre-earthquake income, the highest share of people
whose overall debt has increased are those with a
medium level of income.
Table 4.9: Overall debt - by caste and pre-earthquake income (IRM-4, weighted)
Increased No change Decreased Don’t know/ refused
High caste 50% 39% 10% 1%
Caste Janajati 46% 41% 12% 1%
Low caste 37% 45% 16% 1%
Pre-earthquake income Low 47% 41% 12% 1%
Medium 51% 38% 10% 1%
High 47% 37% 14% 2%
Is more borrowing associated with increased debt?
As expected, more frequent borrowing appears to
have increased overall debt. Those who borrowed
in all three time periods covered in the last three
surveys are 14 percentage points more likely to report
increased debt compared to those who have borrowed
intermittently (Table 4.10).
Table 4.10: Overall debt - by borrowing frequency (IRM-2, IRM-3, IRM-4 panel, unweighted)
Increased No change Decreased Don’t know/ refused
Borrowing frequency Intermittent borrowing 49% 38% 12% 1%
Borrowed in all 3 rounds 63% 24% 12% 0%
Did not borrow during last 3 rounds 8% 73% 8% 10%
Who are the frequent borrowers?
Overall, there are 56% of people who are intermittent
borrowers, having borrowed once or twice in IRM-2,
IRM-3 and IRM-4, and 20% who have borrowed in
all three time periods. People who have sustained
more damage to their house are more likely to
borrow regularly. As shown in Table 4.11, people with
completely destroyed houses are 8 percentage points
more likely to borrow in all three rounds of surveys
than those who suffered major or minor damage
to their house, and almost three times as likely to
borrow in all three rounds as those whose house was
not damaged.
53


Coping Strategies
Table 4.11: Borrowing frequency - by housing damage
(IRM-2, IRM-3, IRM-4 panel, unweighted)
Intermittent borrowing Regular borrowing Did not borrow during last 3 rounds
Completely destroyed 56% 23% 20%
Major damage 58% 15% 27%
Minor damage 52% 15% 33%
No damage 45% 8% 47%
Don’t know 67% 0% 33%
*Note: Intermittent borrowing: borrowed once or twice in IRM-2/IRM-3/IRM-4.
Regular borrowers: Borrowed in all three surveys.
Caste and income. Individuals in both the high
and low caste groups are more likely to be frequent
borrowers compared to Janajatis, who are intermittent
borrowers (Table 4.12). The relationship between
income level and borrowing frequency is more
straightforward. People with low pre-earthquake
income levels are more likely to be regular borrowers.
They are 5 percentage points more likely than those
in the medium income group, and 8 points more
likely than high income individuals, to be regular
borrowers. This trend was first reported in IRM-3
report,28 but the evidence presented here is stronger
and it confirms the ongoing trend of the poorer having
to borrowing increasingly frequently. It suggests that
low income individuals who have borrowed are the
most vulnerable of falling into a debt trap.
Table 4.12: Borrowing frequency - by caste and pre-earthquake income
(IRM-2, IRM-3, IRM-4 panel, unweighted)
Intermittent borrowing Regular borrowing Did not borrow during last 3 rounds
High caste 53% 24% 23%
Janajati 56% 17% 27%
Low caste 55% 25% 20%
Low income 55% 23% 21%
Medium income 54% 18% 27%
High income 55% 15% 30%
Where are people more likely to borrow repeatedly?
Those living in more affected districts and those
in more remote areas are more likely to borrow
repeatedly. As shown in Table 4.13, larger shares of
people in severely (21%) and crisis hit (27%) districts
are borrowing repeatedly, compared to hit with heavy
losses (3%) and hit (19%) districts. Okhaldhunga has
the highest share of people (46%) who have borrowed
in all three rounds of the survey.29 People in more
remote areas are 5 percentage points more likely to
borrow regularly than those in remote areas, and twice
as likely to borrow regularly compared to those in less
remote areas.
28 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 3: September 2016. Quantitative Report. Kathman-
du and Bangkok: The Asia Foundation, p. 60.
29 Okhaldhunga has third highest number of more remote areas after
Solukhumbu and Gorkha. It has the highest level of poverty (80%
mention they are in the low income category).
54


Aid and Recovery in Post-Earthquake Nepal
Table 4.13: Borrowing frequency - by district impact, district and remoteness
(IRM-2, IRM-3, IRM-4 panel, unweighted)
Intermittent borrowing Borrowed in all 3 rounds Did not borrow during last 3 rounds
Severely hit 59% 21% 20%
Dhading 62% 24% 14%
Gorkha 57% 17% 26%
Nuwakot 64% 15% 22%
Ramechhap 61% 28% 12%
Sindhupalchowk 56% 19% 26%
Crisis hit 47% 27% 26%
Bhaktapur 50% 5% 46%
Kathmandu 54% 6% 40%
Okhaldhunga 43% 46% 11%
Hit with heavy losses 52% 3% 45%
Solukhumbu 57% 3% 40%
Lamjung 49% 3% 48%
Hit 56% 19% 25%
Syangja 56% 19% 25%
Less remote 57% 12% 31%
Remote 55% 21% 24%
More remote 55% 26% 19%
Unsuccessful borrowing
Over time, unsuccessful borrowing has increased
slightly. Compared to only 4% of people in IRM-2 who
were unsuccessful in borrowing, 6% of people in IRM-
4 tried to borrow but were unsuccessful. The impact
of the earthquakes has some effect. More people
in severely and crisis hit districts are unsuccessful
compared to those in the bottom two categories of
earthquake impact (Figure 4.14). This may suggest a
higher demand for capital in higher impact districts
to cope with the disaster effect.
However, borrowing success does not appear to as-
sociate well with remoteness. A similar proportion
of people were unsuccessful in borrowing across all
three categories of remoteness in IRM-4 (7%, 6%, 6%).
Figure 4.14: Unsuccessful borrowers - by district impact and remoteness
(IRM-2, IRM-3, IRM-4, weighted)
heavy losses
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
55


Coping Strategies
Caste and income. More higher caste people were
unsuccessful borrowers in IRM-4 (8%) than is the case
for low caste people (6%) or Janajatis (5%). Across
surveys and over time, the proportion of borrowers
who are unsuccessful is increasing for those of high
caste and Janajatis while it is decreasing for the low
caste group, especially when compared to IRM-2
(Figure 4.15). People in the low and high income group
are increasingly likely to be unsuccessful, although
the margin is much higher for the high income group.
Figure 4.15: Unsuccessful borrowers - by caste and pre-earthquake income
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Intention to borrow
The number of people who plan to borrow is increasing.
Thirty-five percent of people in IRM-4 plan to borrow
in the next three months, compared to 28% in IRM-3
and 27% in IRM-2 - Figure 4.16.
More people in the higher impact districts intend to
borrow than was the case before. This trend was also
found in IRM-3.30 However, the difference in bor-
rowing intentions is widening between people living
in severely hit districts and those in other districts.
Whereas 45% of people in severely hit districts plan
to borrow, only 19% in crisis hit districts, 17% in hit
with heavy losses districts and 10% in the hit district
intend to borrow in next three months (Table 4.14).
Ramechhap continues to be the district where the
largest share of people (65%) plan to borrow.
30 The Asia Foundation (2016). Aid and Recovery in Post-Earth-
quake Nepal: Independent Impacts and Recovery Monitoring
Nepal Phase 3: September 2016. Quantitative Report. Kathman-
du and Bangkok: The Asia Foundation, p. 62.
56


Aid and Recovery in Post-Earthquake Nepal
Table 4.14: Share of people who plan to borrow
in the next three months - by district impact and
district (IRM-4, weighted)
Yes No Refused Don’t know
Severely hit 45% 48% 0% 7%
Dhading 54% 35% 0% 10%
Gorkha 33% 60% 0% 7%
Nuwakot 41% 55% 0% 4%
Ramechhap 65% 32% 0% 3%
Sindhupalchowk 40% 53% 1% 7%
Crisis hit 19% 69% 0% 11%
Bhaktapur 24% 62% 0% 14%
Kathmandu 17% 73% 0% 11%
Okhaldhunga 43% 43% 0% 14%
Hit with heavy losses 17% 44% 0% 39%
Lamjung 21% 57% 0% 22%
Solukhumbu 11% 23% 1% 66%
Hit 10% 67% 1% 22%
Syangja 10% 67% 1% 22%
Remoteness. More people in remote and more
remote areas plan to borrow compared to people in
less remote areas (Figure 4.17). People in remote and
more remote areas are 15 percentage points more
likely to say they will borrow than those in less remote
areas.
Gender, widows, disability and housing dam-
age. Intention to borrow is stronger among people
with a disability and those who have suffered greater
damage to their houses. As shown in Figure 4.18, those
with disability are 6 percentage points more likely than
others to say they plan to borrow in the next three
months. Widows are less likely to say they will borrow
than others and the difference by gender is very small
(1%). Borrowing intention correlates highly with the
level of housing damage people experienced. Thir-
ty-nine percent of people whose house was completely
destroyed intend to borrow, compared with 22% who
suffered major damage, 17% with minor damage and
only 7% with no damage. This relationship between
house damage and borrowing intention in IRM-4
reflects the lingering effects of the disasters.
Figure 4.17: Share of people who intend to
borrow in the next three months - by remoteness
(IRM-4, weighted)
Figure 4.18: Share of people who intend to borrow in the next three months - by gender,
widows, disability and housing damage (IRM-4, weighted)
57


Coping Strategies
Caste and income. Low caste people are slightly
more likely than others to say they will borrow in the
next three months. Borrowing intention can be more
accurately predicted by income level (Figure 4.19).
Individuals in the low income group are 11 percentage
points more likely than those in the medium income
group, and almost twice as likely than those with a
higher income, to say they will borrow.
Figure 4.19: Share of people who intend to borrow in the next 3 months - by caste
and pre-earthquake income (IRM-4, weighted)
4.2 Assets sales
Sales of assets have increased and are the most
frequent in more affected districts. While only 4%
of people said they sold assets in IRM-2, and 3% in
IRM-3, 6% now report having sold assets in the last
eight months (Figure 4.20). This rise in asset sales is
largely in the crisis hit districts. Sales of assets remain
highest in the severely hit districts.31
Figure 4.20: Share of people who sold assets to cope with the earthquake impacts -
by district impact (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) H Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
31 District level analysis does not provide robust results because of
the small proportions.
58


Aid and Recovery in Post-Earthquake Nepal
Assets sales in IRM-4 are highest in more remote areas However, the increase in asset sales is sharpest in less
(7%), although only by a margin of 1 percentage point remote areas (Figure 4.21)
compared to remote and less remote areas (both 6%).
Figure 4.21: Share of people who sold assets to cope with the earthquake impacts -
by remoteness (IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
What assets are people selling?
The majority of people who have sold assets in IRM-4
have sold land (43% of those who sold assets) and live-
stock (40%) - Table 4.15. People have also sold gold
(9%) and houses (5%) to cope with the earthquakes’
impacts. Compared to IRM-3, land sales increased by
13 percentage points as a type of asset sold, whereas
livestock sales lowered by 18 percentage points. Land
sales in IRM-4 were the highest in crisis hit districts
(64%), especially in Bhaktapur (95%) and Lamjung
(100%). Gold sales were the highest in Kathmandu
(21%). Land sales in less remote regions were twice
as common as in remote areas and 20 times more
common than in more remote areas. In contrast, live-
stock sales in more remote regions were 23 percentage
points more likely than in remote regions and nearly
six times more likely than in less remote regions.
Table 4.15: Types of assets sold to cope with earthquake impacts amongst those who sold assets -
by district impact, district and remoteness (IRM-4, weighted)
House Land Livestock Household goods Gold Don’t know
Severely hit 2% 14% 77% 5% 2% 0%
Dhading 2% 6% 91% 0% 0% 0%
Gorkha 3% 27% 57% 3% 10% 0%
Nuwakot 0% 32% 68% 0% 0% 0%
Ramechhap 2% 17% 79% 2% 0% 0%
Sindhupalchowk 2% 7% 66% 22% 5% 0%
Crisis hit 7% 64% 12% 1% 16% 0%
Bhaktapur 0% 95% 0% 3% 3% 0%
Kathmandu 11% 63% 5% 0% 21% 0%
Okhaldhunga 0% 5% 92% 0% 3% 0%
Hit with heavy losses 0% 43% 48% 0% 0% 9%
Lamjung 0% 100% 0% 0% 0% 0%
Solukhumbu 0% 7% 79% 0% 0% 14%
Hit 0% 43% 43% 14% 0% 0%
Syangja 0% 43% 43% 14% 0% 0%
59


Coping Strategies
House Land Livestock Household goods Gold Don’t know
All districts 5% 43% 40% 3% 9% 0%
Less remote 9% 61% 14% 2% 14% 0%
Remote 1% 29% 62% 3% 5% 0%
More remote 0% 3% 85% 9% 3% 4%
★Selling assets is a multiple choice response. One asset not mentioned above is selling vehicles,
for which there was only one case in Gorkha
Does repeated borrowing lead to more asset sales?
Borrowing frequency directly correlates with the
likelihood of asset sales. Evidence in the IRM-3 report
suggested that a substantially higher proportion of
people who borrowed in both IRM-3 and IRM-2 had
sold assets compared to those who had borrowed
in one of the rounds only or who had not borrowed
at all.32 Data from the updated panel dataset, which
includes the IRM-4 survey data, confirm the earlier
finding. As before, people who have borrowed more
frequently are more likely to have sold assets. As shown
in Figure 4.22, those who have borrowed repeatedly
since February 2016 (IRM-2) are more than twice as
likely as those who have not borrowed in any of the
three waves to sell assets to cope with earthquake
impacts. They are 3 percentage points more likely than
those who have borrowed intermittently.
Figure 4.22: Number of time periods
borrowed and selling of assets
(IRM-2, IRM-3, IRM-4 household panel, unweighted)
Does where people live or their income
level shape the likelihood of asset sales?
While 6% of people living in their own houses sold
assets in IRM-4, 9% living in shelters on their own
land and 8% living in shelters on others’ land sold
assets to cope with the disaster (Figure 4.23). For
all three groups, these figures represent an increase
since IRM-3.33
32 The Asia Foundation (2016). Aid and Recovery in Post-
Earthquake Nepal: Independent Impacts and Recovery
Monitoring Nepal Phase 3: September 2016. Quantitative
Report. Kathmandu and Bangkok: The Asia Foundation, p. 71.
33 The number of observations for those living in friend’s or
neighbor’s house is less than 1% and hence no meaningful
conclusions can be drawn for these groups.
Figure 4.23: Share of people selling assets - by
where people are living (IRM-3, IRM-4, weighted)
shelter on shelter on other
own land people’s land
Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
60


Aid and Recovery in Post-Earthquake Nepal
Photo: Nayan Pokharel
There is no clear pattern in asset sales in IRM-4 when in IRM-3 (Figure 4.24). Asset sales have increased
disaggregating by income level, as was also the case among all income groups in IRM-4.
Figure 4.24: Share of people selling assets - by
pre-earthquake income (IRM-3, IRM-4, weighted)
Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
61


Coping Strategies
4.3 Remittances
Remittances are becoming more important as a source
of income. Fifteen percent of people in affected areas
say remittances were one of their main income sources
in IRM-4, compared to 9% in IRM-1. However, remit-
tances still tend to be more important in less affected
districts.
Syangja has the highest share of population (34%) who
say remittances is a main income source, although this
figure is lower compared to 41% in IRM-2. In contrast,
Solukhumbu, Lamjung and Dhading districts have
seen notable increases in the number of people relying
on remittances (Table 4.16).
Remittances are less important as an income source
for people in less remote areas (Figure 4.25). They are
most important for people in remote areas and have
also increased sharply in such areas.
Table 4.16: Remittances as a main income
source - by district impact and district
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4)
Severely hit 16% 17% 16%
Dhading 24% 27% 28%
Gorkha 14% 16% 17%
Nuwakot 13% 21% 13%
Ramechhap 14% 12% 12%
Sindhupalchowk 11% 7% 5%
Crisis hit 6% 7% 10%
Bhaktapur 6% 9% 1%
Kathmandu 6% 6% 11%
Okhaldhunga 13% 20% 17%
Hit with heavy losses 19% 25% 28%
Lamjung 23% 27% 30%
Solukhumbu 11% 21% 25%
Hit 41% 39% 34%
Syangja 41% 39% 34%
All districts 13% 14% 15%
Figure 4.25: Remittances as a main income source - by remoteness
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
62


Aid and Recovery in Post-Earthquake Nepal
Remittances are more important as an income source
for those with a high pre-earthquake income and have
grown in importance most for this group (Figure 4.26).
Figure 4.26: Remittances as a main income source - by pre-earthquake income
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
Level of housing damage does not have a strong
effect on the likelihood that people say remittances
are an important income source. Remittances have
become important for more people whose house was
completely destroyed, or damaged (major or minor),
but not for those whose house was not damaged.
However, more people whose house was damaged say
that remittances are important to them compared to
those whose house was destroyed (Figure 4.27).
Figure 4.27: Remittances as a main income source - by housing damage
(IRM-2, IRM-3, IRM-4, weighted)
Feb-Mar 2016 (IRM-2) Sep 2016 (IRM-3) H Apr 2017 (IRM-4)
63


Coping Strategies
Who benefits from remittances?
According to IRM-4, remittances are a major source
of income for 15% of the overall population in affected
districts. Similar to earlier findings,34 IRM-4 responses
indicate that people in the high pre-earthquake income
group benefit the most from remittances. People in the
high income group are 7 percentage points more likely
than those in the low income group, and 8 percentage
points more likely than those with medium income, to
benefit from some form of remittance (from inside or
outside the country) - Table 4.17. Remittances from
outside of Nepal are more common than those from
within the country.
Table 4.17: Share of people receiving remittances - by pre-earthquake income (IRM-4, weighted)
Yes, from inside the country only Yes, from outside the country only Yes, from both inside and outside the country No
Low income 6% 16% 1% 78%
Medium income 5% 13% 3% 79%
High income 3% 24% 2% 71%
As in IRM-3, level of housing damage does not seem
to have any association with the likelihood of receiving
remittances - Table 4.18. For instance, people with
minor damage to their houses are more likely to
receive remittances (30%) than those with major
damage (27%) or complete destruction (23%).
Table 4.18: Share of people receiving remittances - by housing damage (IRM-4, weighted)
Yes, from inside the country only Yes, from outside the country only Yes, from both inside and outside the country No
Completely destroyed 5% 16% 2% 77%
Major damage 5% 19% 4% 73%
Minor damage 4% 24% 2% 70%
Not damaged 1% 15% 2% 81%
Where people live is also not correlated with the likeli-
hood of them receiving remittances in IRM-4. People
living in their own houses are comparable in their
likelihood of receiving some form of remittance (23%)
to those living in shelters on their own land (25%)
or on other’s land (24%), but more likely than those
living in a neighbor’s house (15%) - Table 4.19. The
highest share of people receiving some form of remit-
tance are individuals living in rented accommodation
(27%), who are nearly 2% of the affected population
according to IRM-4.
Table 4.19: Share of people receiving remittances - by where people are living (IRM-4, weighted)
Yes, from inside the country only Yes, from outside the country only Yes, from both inside and outside the country No
Own house 3% 18% 2% 77%
Neighbor’s house 6% 9% 0% 85%
Friend’s house* 0% 27% 0% 73%
34 The Asia Foundation (2016). Aid and Recovery in Post-
Earthquake Nepal: Independent Impacts and Recovery
Monitoring Nepal Phase 3: September 2016. Quantitative
Report. Kathmandu and Bangkok: The Asia Foundation, pp.
73-74.
64


Aid and Recovery in Post-Earthquake Nepal
Yes, from inside the country only Yes, from outside the country only Yes, from both inside and outside the country No
Self-constructed shelter on own land 8% 16% 1% 75%
Self-constructed shelter on other people’s land 3% 19% 2% 76%
Self-constructed shelter on public land 14% 0% 0% 86%
Community shelter* 0% 0% 0% 100%
Rent 1% 18% 8% 73%
Relative’s house* 0% 100% 0% 0%
*Less than 1% people in IRM-4 are living in friend’s house, community shelters ora relative’s house
Improvements of income sources is also not correlated
with the likelihood of receiving remittances, which was
also the case in IRM-3. Twenty-two percent of people
who say that their income sources have not improved
are receiving some form of remittance, similar to those
whose income sources have improved (21%).
4.4 Migration
Most people say that the level of out-migration from
their community has stayed the same compared to
before the earthquakes. However, more people say
that migration has increased than decreased. Overall,
65% of people say migration levels have remained the
same, 20% say they have decreased, and 4% say levels
have decreased (Table 4.20).
There is no clear pattern in reported migration by
the level of earthquake impact. For example, some
severely hit districts have reportedly seen increases
in out-migration from communities (27% in Dhading
say it has increased) but other have low reports of
increases (for example, in Nuwakot where 4% say
it has increased). Lesser affected districts such as
Kathmandu (29% report an increase) and Syangja
(18%) have particularly high proportions of people
reporting increases in out-migration. More people in
Nuwakot say that out-migration has decreased than
say it has increased.
Table 4.20: Change in number of people migrating from respondents’ communities compared
to before the earthquake - by district impact and district (IRM-4, weighted)
Increased Stayed the same Decreased Don’t know/refused
Severely hit 14% 80% 3% 3%
Dhading 27% 67% 1% 6%
Gorkha 15% 77% 4% 4%
Nuwakot 4% 82% 9% 5%
Ramechhap 6% 93% 0% 1%
Sindhupalchowk 11% 86% 2% 0%
Crisis hit 25% 52% 4% 19%
Bhaktapur 11% 82% 5% 1%
Kathmandu 29% 44% 4% 23%
Okhaldhunga 4% 94% 1% 0%
Hit with heavy losses 10% 85% 4% 1%
Solukhumbu 7% 90% 1% 2%
Lamjung 12% 82% 6% 0%
Hit 18% 78% 2% 3%
Syangja 18% 78% 2% 3%
All districts 20% 65% 4% 11%
65


Coping Strategies
Increases in reported out-migration are greater in less
remote areas (Figure 4.28).
Figure 4.28: Change in number of people migrating from respondents’ communities compared
to before the earthquake - by remoteness (IRM-4, weighted)
Increased Stayed the same H Decreased
Migration plans in the next year
Only 3% said that they, or someone in their family,
planned to migrate in the next year. A majority of
those who have a plan to migrate are from severely hit
districts (61%) or crisis hit ones (28%). These results
suggest a migration pattern following the disaster.
After the earthquake, migration increased as many
people in less remote areas moved to more remote
areas, possibly to help families in need. But two years
after the disaster, there are more people in higher
impact districts and remote areas that are looking to
migrate elsewhere.
Migrating within the country or abroad?
A higher proportion of people from remote and less
remote areas plan to migrate abroad, but a majority
of people in more remote areas who plan to migrate in
the next 12 months plan to migrate within the country.
There is no consistent relationship between income
level, earthquake impact or level of debt and where
people want to migrate to.
66


Aid and Recovery in Post-Earthquake Nepal
Photo: Ishwari Bhattarai
This chapter reviews aid received since the earthquakes
struck and how aid flows have changed over time. Aid
coverage, types received, and who is providing aid are
discussed. Needs on the ground and how they have
changed over time are also examined. The chapter
looks at changes in satisfaction with aid providers and
perceptions of fairness of aid distribution.
Key Findings
Aid coverage
• In April 2017, 40% of those in earthquake-
affected areas said they had received aid since the
end of monsoon. Since IRM-3, the share getting
aid has gone up by 25 percentage points. This is
largely due to the distribution of the first tranche
of the government’s housing grant.
• Recent aid distribution has been concentrated in
the districts that were severely hit and crisis hit.
• People in less remote areas are the least likely to
have received aid. Aid received decreases with
income, and those belonging to higher castes are
less likely to have received aid. Similar shares of
men and women, and those with and without a
disability, have received aid.
• Few said no aid was required, either at the time
of the April 2017 survey or in the near future. In
the severely hit districts, less than 5% said that
no aid was required.
• The government has been the foremost aid
provider since the earthquake, and is almost the
sole provider of material aid since winter 2016. In
recent months, cash has been the most common
form of assistance.
• Those who received cash assistance from the gov-
ernment have received on average NPR 56,845 to
date; those who received it from non-governmen-
tal sources have got NPR 13,082.
Needs in earthquake-affected areas
• Seven in 10 mention cash as a current need and
over six in 10 mention it as a future need. Recon-
struction materials, the next most frequently cited
item, is mentioned by only 30% as a current need
and 18% as a future need.
• Mention of cash as a need has been growing
steadily: 38% said it was a current need in IRM-1
while 64% said it will be needed in the near fu-
ture in IRM-4. Nearly everyone in the severely
hit districts mentioned cash when asked about
immediate and future needs. Fewer people in less
affected areas mentioned it. The stated need for
cash is higher in more remote areas.
67


Earthquake Aid
• People with lower incomes are more likely than
those with higher incomes to have mentioned any
item as a need.
Satisfaction with aid distribution
• Satisfaction with most aid providers was highest
in the period right after the earthquake. It
plunged after February 2016 and has stayed at
similar levels since then.
• The security forces—the army, the armed police
force and the police—still get high ratings for
their work in earthquake-affected areas, though
their direct involvement was limited to the early
response period.
• People express the lowest levels of satisfaction
with local political parties, religious groups
and private businesses for their involvement in
earthquake relief since the end of winter.
• Those in the severely hit districts have been the
most likely to think that aid distribution has been
fair in all four surveys and the share of people in
these districts who believe distribution has been
fair has remained stable. Less affected areas have
seen more fluctuation in opinions as to whether
aid distribution has been fair.
• People with higher incomes are less likely than
those with lower incomes to think that aid
distribution has been fair.
• Most people who think aid distribution has been
unfair believe that those belonging to lower castes
are unable to receive aid equally and according
to their needs.
• Those belonging to lower castes think they are
more likely to be treated unfairly by a wide
margin: 64% of those belonging to lower castes
believe they are most likely to be treated badly,
compared to 39% of those of high caste and 36%
of Janajatis.
Aid communication
• Over seven in 10 mentioned neighbors as their
prime source of aid information in both IRM-3
and IRM-4. People with higher incomes, and
those belonging to higher castes, are less likely
than others to say that neighbors are their top
source of information on aid.
• Just over half are satisfied with how local commu-
nity organizations, the army, the police and the
armed police force have communicated about aid.
Over half are dissatisfied with how local political
parties have informed them about aid.
• Compared to IRM-3, people are more likely to be
unsure about how to rate communication with
aid providers.
• People think that communication with most
aid providers is either bad or okay; few say that
communication with aid providers is good.
5.1 Aid coverage
Changes in aid coverage
Two years on from the earthquakes, 40% of the
people in earthquake-affected areas say that they have
received aid since the end of the last monsoon. In the
early months after the earthquakes almost everyone
said they received some type of aid (96% in IRM-1) -
Figure 5.1. One year after the quakes about half said
they received aid in the period between IRM-1 and
IRM-2 (54% in IRM-2). By September 2016, there
was a huge drop in the share saying they received aid
(15% in IRM-3).35 This jump in the number of people
receiving aid is largely due to the completion of the
distribution of the first tranche of the housing grant
from the National Reconstruction Authority (see
Chapter 6). Since last winter, aid coverage has been
concentrated in the severely hit (where 81% received
aid) and crisis hit (25%) districts.
35 The Asia Foundation (2017). Aid and Recovery in Post-
Earthquake Nepal: Independent Impacts and Recovery
Monitoring Nepal Phase 3: September 2016. Quantitative
Report. Kathmandu and Bangkok: The Asia Foundation
pp. 81-82.
68


Aid and Recovery in Post-Earthquake Nepal
Figure 5.1: Share of people receiving some type of aid - by district impact
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) H Sep 2016 (IRM-3)
Mar-Feb 2016 (IRM-2) H Apr 2017 (IRM-4)
People in Sindhupalchowk (89%), Ramechhap (80%)
and Nuwakot (84%) were the most likely to have
received aid since the end of the monsoon (Figure 5.2).
Though within the same earthquake impact category,
people in Kathmandu (18%) were much less likely
to get aid than in Bhaktapur (60%) or Okhaldhunga
(51%). Almost no-one received aid in Lamjung and
Solukhumbu and just 4% in Syangja. Of the districts
surveyed, these were the only ones where the first
tranche of the government’s housing grant had not
been disbursed at the time of the survey.
Figure 5.2: Share of people receiving some type of aid since the end of winter 2016 -
by district (IRM-4, weighted)
However, the recent increase in aid coverage has
not been seen everywhere. Aid coverage declined in
hit with heavy losses and hit districts in the same
period (Table 5.1). Increases in aid coverage between
IRM-3 and IRM-4 were sharpest in Nuwakot (up 69
points), Dhading (up 66 points) and Bhaktapur (up
69


Earthquake Aid
6o points). Previously, there was a sharp decline in
the share receiving aid in all districts between IRM-2
and IRM-3. Between IRM-1 and IRM-2, there was
not much change in aid distribution in the severely
hit districts, but in other areas the share receiving aid
declined sharply, except in Solukhumbu.
Table 5.1: Change in aid coverage - by district and district impact (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Feb- Mar 2016 (IRM-2) Sep 2016 (IRM-3) Apr 2017 (IRM-4) Change in coverage between IRM-1 and IRM-2 (percentage points) Change in coverage between IRM-2 and IRM-3 (percentage points) Change in coverage between IRM-3 and IRM-4 (percentage points)
Severely hit 100% 98% 26% 81% -2% -72% 55%
Dhading 100% 97% 7% 73% -3% -90% 66%
Gorkha 100% 97% 56% 79% -3% -41 % 23%
Nuwakot 100% 99% 15% 84% -1% -84% 69%
Ramechhap 100% 97% 21% 80% -3% -76% 59%
Sindhupalchowk 100% 100% 32% 89% 0% -68% 57%
Crisis hit 92% 30% 11% 25% -62% -19% 14%
Bhaktapur 100% 55% 0% 60% -45% -55% 60%
Kathmandu 91% 23% 11% 18% -68% -12% 7%
Okhaldhunga 100% 76% 34% 51% -24% -42% 17%
Hit with heavy losses 100% 65% 6% 0% -35% -59% -6%
Lamjung 100% 47% 0% 0% -53% -47% 0%
Solukhumbu 100% 95% 16% 0% -5% -79% -16%
Hit 100% 30% 5% 4% -70% -25% -1%
Syangja 100% 30% 5% 4% -70% -25% -1%
All districts 96% 54% 15% 40% -42% -39% 25%
Aid coverage among different population groups
Remoteness. Whereas aid coverage was extremely
widespread everywhere in IRM-1, from then on more
remote areas started to receive more aid compared to
other areas. Even in IRM-3, when aid coverage was
at its lowest, people in more remote areas were twice
as likely to have received aid compared to areas that
were less remote or remote (Figure 5.3).
Figure 5.3: Proportion who received aid - by remoteness (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Less remote H Remote H More remote
70


Aid and Recovery in Post-Earthquake Nepal
Gender. In all four survey rounds, men and women
have been equally likely to report that their household
received some form of aid. In IRM-4, 39% of men and
40% of women said they received aid.
Widows. Forty-three percent of widows said they
received aid in IRM-4. Just over half of widows got
aid in IRM-1 and IRM-2 (53% in each round), while
only 15% received aid in IRM-3.
Caste. People belonging to higher castes continue
to be less likely to have received aid compared to
Janajatis and lower castes (Figure 5.4).
Figure 5.4: Proportion who received aid - by caste (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
High castes H Janajatis H Low castes
Disability. Having a disability has not affected Income. As income rises, the likelihood of having
whether or not someone receives aid. In IRM-4,41% of received aid decreases sharply (Figure 5.5). This pat-
those with a disability and 39% without one reported tern has also been evident in all earlier IRM surveys,
having received aid.
Figure 5.5: Proportion who received aid - by pre-earthquake income
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Low income Middle income H High income
71


Earthquake Aid
Do people still require assistance?
Though aid provision, besides the government cash
grants (examined in Chapter 6), has almost come to
a halt, needs remain. However, the share of people
saying no relief is or will be needed has been shrinking.
In IRM-4, just 3% in the severely hit districts say no
relief is needed, compared to about three in 10 in
areas less affected by the earthquake. In the severely
hit districts, less than 5% have said no relief is/will be
needed in all four surveys. In the crisis hit districts,
the share saying no aid is necessary declined 32 points
between June 2015 and April 2017, and is the same
for April 2017 and what is anticipated for six months
later. In the hit with heavy losses districts, only about
one-third of people said that there was no need for aid
at present (Table 5.2).
Almost no-one in the severely hit districts said that
aid is/will not be needed. Of the crisis hit districts,
Okhaldhunga residents are far less likely than those
in Kathmandu and Bhaktapur to have said no relief
is required at any point. Lamjung is the only district
that has shown improvements: compared to IRM-3,
people in Lamjung were more likely to say no aid is
required at present in IRM-4. Also, 64% said they do
not need relief in the future.
Despite more aid going to more remote areas, the
share saying no relief is needed, either at present or in
the future, declines sharply with remoteness. Among
those in less remote areas, 33% said that aid is not
required either at present or in the future. In remote
areas, 18% said they do not need aid at present and
15% said that it would not be required in the future. In
more remote areas, just 2% said no aid was required
at present, and 1% said aid would not be required in
the future.
Table 5.2: Share saying they do not need aid either now or in the future - by district impact and district
(IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Jun 2015 (IRM-1) Dec 2015 projected (IRM-1) Feb-Mar 2016 (IRM-2) Sep 2016 projected (IRM-2) Sep 2016 (IRM-3) Mar 2017 projected (IRM-3) Apr 2017 (IRM-4) Oct 2017 projected (IRM-4)
Severely hit 1% 1% 2% 6% 2% 2% 3% 4%
Dhading 1% 1% 2% 9% 3% 3% 3% 3%
Gorkha 3% 3% 5% 6% 4% 4% 3% 3%
Nuwakot 0% 0% 0% 0% 1% 1% 3% 3%
Ramechhap 0% 2% 1% 1% 0% 0% 0% 1%
Sindhupalchowk 1% 2% 1% 14% 1% 0% 2% 7%
Crisis hit 65% 74% 60% 60% 42% 42% 33% 33%
Bhaktapur 37% 39% 39% 39% 35% 35% 28% 27%
Kathmandu 73% 83% 66% 66% 45% 45% 36% 36%
Okhaldhunga 8% 24% 8% 10% 7% 7% 9% 9%
Hit with heavy losses 37% 50% 34% 48% 29% 29% 33% 41%
Lamjung 56% 70% 46% 56% 40% 40% 51% 64%
Solukhumbu 3% 13% 13% 34% 9% 9% 1% 1%
Hit 55% 64% 55% 58% 74% 74% 27% 31%
Syangja 55% 64% 55% 58% 74% 74% 27% 31%
All districts 42% 49% 42% 45% 30% 30% 23% 24%
Who is providing aid?
The government has been the top aid provider since the
earthquakes struck (Figure 5.6). In each survey, those
who said they received aid were asked where they got
the aid from. In IRM-4, almost everyone who received
aid (96%) received it from the government. Even in early
surveys, at least six in 10 named the government as the
source of earthquake aid. NGOs, INGOs, individual do-
nations, and the Red Cross have been other top sources
of aid. NGOs were a consistent aid provider until IRM-4,
with about one-quarter of aid recipients naming them
as the source of aid in earlier surveys. INGOs were most
active in the period between June 2015 and February
2016 (31% IRM-3). Individual donations accounted for
15% of aid received in the early response period (IRM-
1), but it shrunk to 7% by IRM-2 and was only 1% of the
aid provided in both IRM-3 and IRM-4. The Red Cross
was also named as an aid provider in IRM-1 (12%) and
IRM-2 (15%), but fewer mentioned it in later surveys.
72


Aid and Recovery in Post-Earthquake Nepal
Figure 5.6: Source of aid amongst those who received aid (IRM-1, IRM-2, IRM-3, IRM-4, weighted)
Apr 2017 (IRM-4)
Sep 2016 (IRM-3)
H Feb-Mar 2016 (IRM-2)
Jun 2015 (IRM-1)
Among those who received aid since the last monsoon,
cash (39%) is the most cited aid item received. Other
items such as tarps (4%), food (3%) and corrugated
iron sheets (1%) are mentioned far less frequently.
Though cash is the most needed item, and now the
main type of aid provided, 20% of people who received
cash from the government say they have not yet used it.
How much money have people received to date?
IRM-4 also looked at how much cash assistance in total
people had received since the earthquakes, either from
the government or from non-governmental sources.
By April 2017, 55% of people in earthquake-affected
areas had received cash from the government and 6%
from non-governmental sources (Figure 5.7). In earlier
surveys, a slightly smaller share said they had received
aid from the government and a similar share said they
had received cash from non-governmental sources.
This suggests that most government cash has been
targeted at those who previously received it, although
there are some new beneficiaries, while there has been
little non-government cash going to new people.36
Figure 5.7: Share receiving cash from the
government and non-governmental sources
(IRM-2, IRM-3, IRM-4, weighted)
Government H Non-governmental sources
36 The reduction in people who say they have received cash from
non-government agencies over time suggests that people are
forgetting about earlier assistance from non-government
providers.
73


Earthquake Aid
Among those who have received cash, those in earth-
quake-affected areas say they have received on average
NPR 56,845 from the government since the earth-
quakes struck and NPR 13,082 from non-governmental
sources. The largest amount from the government was
received between November 2016 and April 2017, which
was the period when many areas got the first tranche of
the National Reconstruction Authority’s (NRA) housing
grant. The government also provided two smaller cash
grants of NPR 15,000 and NPR 10,000 to support the
construction on temporary shelters and for winter re-
lief. Most cash grants from non-governmental sources
were given just after the earthquakes, and little has been
distributed more recently.37 * * 37
In eight of the eleven surveyed districts—all except
Lamjung, Solukhumbu and Syangja—the first tranche
of the cash grant (NPR 50,000) was provided before
April 2017 when the survey was conducted. This
is reflected in the cumulative amounts people say
they have received from the government in the nine
districts, which all exceed NPR 50,000 with the
exception of Okhaldhunga. In contrast, government
cash recipients in the districts where the housing grant
had not yet been rolled out have received far less with
the exception of Solukhumbu where people say they
have received more (Table 5.3).
Since the earthquakes, the largest amount of non-
governmental cash by far has gone to Solukhumbu
according to IRM-3 data, although reported amounts
received have declined in IRM-4 for reasons we cannot
explain.
Table 5.3: Average cash amount received to date (NPR) from the government and non-governmental
sources - by district impact and district (IRM-2, IRM-3, IRM-4, weighted)
Government Non-governmental
IRM-2 IRM-3 IRM-4 IRM-2 IRM-3 IRM-4
Mean Mean Mean Mean Mean Mean
Severely hit 24,245 31,511 68,545 11,901 14,586 12,98338
Dhading 24,552 28,433 65,922 11,908 11,425 5,951
Gorkha 17,342 35,738 66,569 12,006 21,433 21,798
Nuwakot 29,924 32,685 73,288 9,790 11,554 13,459
Ramechhap 24,845 32,759 66,835 - 4,000 7,000
Sindhupalchowk 24,354 28,911 70,197 12,214 12,016 11,136
Crisis hit 24,569 22,528 50,388 11,211 11,853 13,368
Bhaktapur 24,224 24,999 56,669 11,225 8,914 21,823
Kathmandu 26,749 22,687 56,669 11,049 24,758 13,521
Okhaldhunga 16,708 18,266 48,932 14,031 5,885 5,217
Hit with heavy losses 15,923 15,015 15,083 14,490 29,030 13,275
Lamjung 23,000 23,811 23,409 14,259 12,368 13,182
Solukhumbu 12,420 10,949 49,140 16,700 109,071 14,000
Hit 8,203 7,766 8,334 3,821 5,400 11,986
Syangja 8,203 7,766 8,334 3,821 5,400 11,986
All districts 23,273 26,586 56,845 11,553 14,194 13,082
37 There are restrictions to how non-governmental agencies can give
cash aid. In the early response period, the government followed
a one-door policy, whereby all donations had to go through the
government. In later periods as well, there are limitations to how
non-governmental agencies can give cash grants. Major donors
have been the Red Cross, UNICEF, which has a program focused
on school children, and JICA, who worked on housing grants
following similar guidelines as the NRA.
38 Reported reductions in cash received are likely due to problems
with recall.
74


Aid and Recovery in Post-Earthquake Nepal
5.2 People’s needs in
earthquake-affected areas
What are current needs?
Cash (69%) and items to reconstruct people’s houses
(30%) are the top current needs stated by respondents.
Other items mentioned include clean drinking water
(9%), rice, wheat and maize (7%), and corrugated iron
sheets, clean water for household purposes and farm
implements (6% each).39
Strong majorities express a need for cash across all
districts, but it is highest in the severely hit districts
(89%) - Table 5.4. For other most commonly cited
needs, too, higher shares in the severely hit districts
mention them compared to less impacted districts.
Nearly everyone in Dhading (97%), Ramechhap (96%)
and Solukhumbu (94%) said they need cash at present.
Items to reconstruct houses are mentioned most often
in Nuwakot (65%), Ramechhap and Sindhupalchowk
(42% each). Clean drinking water is a top need in Nu-
wakot (21%), Ramechhap (18%) and Sindhupalchowk
(15%). Water for household purposes is more com-
monly cited as a priority need in Dhading (15%) and
Sindhupalchowk (13%). Four in 10 Nuwakot residents
mention farm implements and two in 10 Okhaldhunga
residents mention corrugated iron sheets.
Remoteness. The need for cash increases sharply
with remoteness. Those in less remote areas (19%)
are far less likely to mention items to reconstruct their
house than people living in more remote and remote
areas (40% and 45%, respectively).
Table 5.4: Most mentioned current needs - by district impact and district (IRM-4, weighted)
Cash Items to reconstruct house Clean drinking water Rice, wheat, maize Farm implements Corrugated iron sheets Clean water for household purposes
Severely hit 89% 46% 13% 14% 13% 11% 10%
Dhading 97% 27% 7% 3% 0% 6% 15%
Gorkha 86% 41% 9% 25% 9% 9% 7%
Nuwakot 85% 65% 21% 15% 41% 15% 8%
Ramechhap 96% 52% 18% 8% 4% 9% 4%
Sindhupalchowk 81% 52% 15% 18% 14% 16% 13%
Crisis hit 59% 22% 8% 2% 2% 2% 5%
Bhaktapur 65% 13% 6% 5% 4% 1% 1%
Kathmandu 56% 22% 8% 1% 1% 0% 6%
Okhaldhunga 78% 37% 7% 14% 8% 23% 2%
Hit with heavy losses 64% 33% 1% 6% 8% 15% 1%
Lamjung 47% 31% 0% 4% 11% 13% 0%
Solukhumbu 94% 37% 3% 10% 3% 18% 3%
Hit 69% 19% 2% 9% 5% 6% 1%
Syangja 69% 19% 2% 9% 5% 6% 1%
All districts 69% 30% 9% 7% 6% 6% 6%
Income and disability. The stated need for all
items declines with income. For instance, 84% with
a low pre-earthquake income say they require cash
compared to 67% in the medium income category and
56% in the high income one (Table 5.5). Needs among
those with and without a disability are similar.
39 Respondents could mention up to three different items. Therefore,
percentages add up to more than 100%.
75


Earthquake Aid
Table 5.5: Most mentioned current needs - by pre-earthquake income (IRM-4, weighted)
Cash Items to reconstruct house Clean drinking water Rice, wheat, maize Farm implements Corrugated iron sheets Clean water for household purposes No need of relief
Low income 84% 39% 10% 14% 13% 11% 10% 6%
Medium income 67% 29% 9% 4% 4% 4% 7% 25%
High income 56% 23% 8% 2% 2% 2% 3% 37%
Gender and caste. Stated needs among men and belonging to lower castes are most likely to mention
women do not differ. Widows are more likely than a need for all items, followed by Janajatis and those
others to say they require cash (75% to 69%). Those in the higher caste group (Table 5.6).
Table 5.6: Most mentioned current needs - by gender, widows and caste (IRM-4, weighted)
Cash Items to reconstruct house Clean drinking water Rice, wheat, maize Farm implements Corrugated iron sheets Clean water for household purposes No need of relief
Female 70% 31% 9% 7% 6% 6% 6% 22%
Male 69% 29% 9% 7% 6% 6% 7% 24%
Widows 75% 29% 5% 11% 6% 6% 6% 17%
High caste 65% 26% 7% 4% 5% 5% 6% 27%
Janajati 71% 32% 9% 8% 7% 7% 6% 21%
Low caste 78% 37% 8% 13% 8% 9% 7% 15%
Where people live. Those in temporary shelters, or
living in a neighbor’s house, are far more likely than
those in their home to say they need cash (Table 5.7).
Those still living in self-constructed shelters on other
people’s land (64%) are the most likely to mention
items to reconstruct house. Other items are also
mentioned most often by those in self-constructed
shelters compared to other types of housing. Three in
10 living in their own house (30%) and renting (28%)
say they do not need any relief items while almost no-
one (between 0% and 1%) living elsewhere says this.
Table 5.7: Most mentioned current needs - by where people are living (IRM-4, weighted)
Cash Items to reconstruct house Clean drinking water Rice, wheat, maize Farm implements Corrugated iron sheets Clean water for household purposes No need of relief
Own house 63% 22% 8% 5% 4% 4% 6% 30%
Neighbor’s house* 97% 24% 4% 5% 5% 15% 9% 3%
Self-constructed shelter on own land 90% 54% 12% 12% 13% 13% 9% 1%
Self-constructed shelter on other people’s land 87% 64% 8% 17% 8% 11% 3% 1%
Rent* 71% 31% 3% 3% 3% 1% 0% 28%
*Small sample sizes. Renters make up 1% and those living with neighbors slightly less than 1% of surveyed population
76