Vulnerable Populations Survey Sri Lanka 2011
42
Vulnerable Populations Survey Sri Lanka 2011 Colombo, Sri Lanka
Transcript of Vulnerable Populations Survey Sri Lanka 2011
Vulnerable Populations Survey
Sri Lanka 2011
Colombo, Sri Lanka
1
EXECUTIVE SUMMARY
Sri Lanka is a multi‐ethnic, multi‐religious society with four major religions: Buddhism, Hinduism, Islam, and
Roman Catholicism. As the country emerges from decades of war, new ethno‐religious dynamics are
emerging which have implications for peace and coexistence between communities. This survey is meant
to help uncover some of those dynamics, especially as they relate to Sri Lanka’s Muslims. In Sri Lanka,
ethnicity and religion are intertwined. Buddhists are Sinhalese; Hindus are Tamils; Muslims comprise a
separate ethnic category and are still referred to as ‘Moors’, especially among older Sri Lankans. While
Muslims’ mother‐tongue is Tamil, they are not Tamils. Christians are either Sinhalese or Tamil, and the
overwhelming majority is Roman Catholic.
The concerns of Sri Lanka’s Muslim community have largely been left out of the “national question” due to
the dominance of the Sinhalese‐Tamil political and social conflict. However, the growing demographic
strength of the Muslim community is likely to interfere with this binary narrative about peace, conflict, and
coexistence, and in the future, lead to a multiplicity of national questions. The Muslim community’s size in
relation to other communities’ will be brought into sharper relief in late 2012 with the conclusion of the
first all‐island household census since 1981.
Within this complex and evolving landscape, The Asia Foundation conducted the Vulnerable Populations in
Sri Lanka Survey in late 2011 to gain a more grounded understanding of people’s perceptions of religious
beliefs and practices, influence of religious leaders, inter‐religious relations, and the extent to which it
contributes to tension and violence particularly as it relates to Sri Lanka’s Muslim population. The survey
revealed the following key findings:
A large plurality of Sri Lankans believe the country is moving in the right direction, however religious
minorities are less positive than Buddhists. Overall, 43 percent of Sri Lankans believe the country is
moving in the right direction while just 15 percent see it moving in the wrong direction. An additional
23 percent feel some things are going in the right direction and some things the wrong direction.
Differences in perception can be seen between religious groups: 49 percent of Buddhists feel the
country is moving in the right direction compared to 38 percent of Muslims and just 22 percent of
Hindus. Nearly twice as many Muslims (21%) and Hindus (23%) as Buddhists (12%) believe the country
is moving in the wrong direction. Infrastructure improvements and the end of the war are seen are the
main reasons for those who feel the country is moving in the right direction. Increases in the price of
goods and an increase in corruption and bribery are the main reasons for those who feel the country is
moving in the wrong direction.
Most Sri Lankans feel that the conclusion of the war has not brought an end to ethnic conflict in the
country but that it has been reduced. Over half (53%) of Sri Lankans feel that the end of the war has
not ended ethnic conflict in the country but that it has been reduced, while 32 percent believe that the
conclusion of the war has indeed brought an end to ethnic conflict in Sri Lanka. Seven percent believe
the end of the war has not changed anything.
Sri Lankans overwhelmingly perceive their society as becoming significantly more religious.
Nationally, 63 percent believe people are much more religious now than 5 years ago. More Buddhists
(70%) and Muslims (53%) say they are “much more religious than other either Hindus (39%) or
Catholics (44%).
2
Observance of core Islamic practices is very high among Sri Lankan Muslims, and certain practices
have increased in recent years. More than 90 percent of Muslims fast during Ramadan and pray five
times a day. Three quarters of Muslim women in Sri Lanka now cover, representing a marginal increase
(+12%) compared to five years ago. However, most Muslim women only cover partially, and primarily
as a matter of personal choice. The majority of Muslims (51%) prefer to consult the Holy Quran to
determine whether or not a certain action is against the teachings of Islam, while one quarter would
consult their local imam.
A majority of Sri Lankan Muslims believe that at least to a certain extent traditions and practices of
Islam in Sri Lanka are different from other countries with Muslim populations. While not exhibiting a
strong opinion, Sri Lankan Muslims generally feel that their brand of Islam is different than other
countries with 43 percent saying they “somewhat agree” with the notion and 18 percent strongly
agreeing. Only 18 percent of Muslims disagree with the idea that Sri Lanka’s brand of Islam differs from
other countries.
Muslims generally view foreign influence on their religious teachings positively, Buddhists view it
negatively. A plurality of Muslims (38%) believe foreign influence on their religious teachings is a good
thing, compared to just 9 percent of Buddhists. For Muslims, the countries perceived as having the
greatest influence are Saudi Arabia (39%), Pakistan (18%), and India (18%). Buddhists cite India (27%)
and the United States (20%).
Local religious leaders are by far the most respected leaders in the country for Sri Lankans of all
faiths. Nationally, 80 percent say local religious leaders are ‘very well respected’, compared to 32
percent for GN officials and just 12 percent for politicians. While the majority of Sri Lankans (66%)
believe religious leaders should not be involved in politics, Muslims are relatively split on the issue,
with just under half (42%) believing religious leaders should have some role to play in politics. Among
Muslims, 52 percent consider the Maulvi to be their local leader while 43 percent consider their Imam.
Rather than indicating internal disagreement, this split in views among Muslims likely reflects the fact
that in some communities Maulvis—who are primarily religious scholars—often also play the role of
Imam, leading prayers and providing spiritual guidance.
Sri Lankans unanimously believe religious leaders should play a central role in community level
development, however few are currently involved. While Sri Lankans believe local religious leaders
should stay out of politics, a staggering 84 percent think that religious leaders should be involved in the
development of neighborhood amenities. A further 81 percent believe that if religious leaders are not
consulted on the problems facing an area, political leaders will make mistakes. Despite this sentiment,
only a quarter (26%) say religious leaders in their area are actually participating in local development
on a regular basis, and among Muslims the rate is even lower (17 percent).
Religious education is increasingly important for Sri Lankans of all faiths. Nine out of ten Muslims and
Buddhists say that religious education is more important today than five years ago. Nearly three‐
quarters of Muslims respondents or their family are currently attending or have attended a religious
school, and the rate is even higher among Buddhists (87%). Religious education among Sri Lankan
Muslims is almost always pursued domestically with only a small proportion (4%) indicating they have
3
studied abroad, the vast majority doing so in Saudi Arabia. If presented with an opportunity to pursue
religious studies abroad, Muslims overwhelmingly would prefer to do so in Saudi Arabia.
Religious minorities, including Muslims, are more likely than Buddhists to perceive discrimination by
the government. Overall, Sri Lankans believe that the rights of minorities have improved at least to
some degree compared to five years ago (47% much better, 32% somewhat better), however Buddhists
believe improvement has been much more substantial than minorities themselves who are more likely
to feel only minor or no change has taken place. Nine in ten Buddhists (90%) believe that people of all
religions are treated equally in the government job market, compared to 66 percent of Muslims and 45
percent of Hindus. Over two‐thirds (66%) of Sri Lankans feel religious leaders are most active in working
to ensure that the rights of religious minorities are safeguarded, while only 2% feel local officials or
national politicians are the most active. Roughly twice as many Hindus (50%), Muslims (41%), and
Catholics (41%) than Buddhists (21%) strongly believe that government should enact special initiatives
to protect the land and assets of minorities.
A majority feel the country is vulnerable to violence stemming from extremist religious views. When
it comes to violence in the country, while Sri Lankans perceive the majority of it to be related to either
criminal activities or political operations, 15 percent cite religious or ethnically inspired violence and
this rate increases to a quarter among Muslims and Hindus. Furthermore, 51 percent of all
respondents feel the country is vulnerable to violence related to extremist religious views. Muslims are
divided regarding the existence of intra‐religious conflict in their community. While relatively few
Muslims outright agree (12%) with the assertion that there are differences of opinion among Muslims
on matters of religion that are leading to conflict, less than half (45%) disagree with the assertion and a
further 43 percent are neutral or do not answer.
Ten geographic areas of Sri Lanka consistently show high levels of inter‐religious tension and support
for the use of violence to achieve religious or political objectives: Mannar Island, Akurana, Kaluthara‐
Beruwala, Puttalam, Southern Vavuniya, Trincomalee, Ampara, Batticaloa, southern Nuwara Eliya,
Badulla. In these areas the survey revealed:
o Intolerance is high for certain forms of religious expression. Sixty‐four percent of Sri Lankans
believe attempts to convert people from one religion to another should not be tolerated, a
perception that is relatively consistent across the four surveyed religious groups. Erecting
religious statues or places of worship in public places continues to be divisive between and
within all religious groups. Forty‐two percent of Buddhists believe religious statues or places of
worship should be allowed in public places compared to 30 percent of Muslims and just 23
percent of Hindus. At the same time, nearly equal proportions within each group believe
statues or places of worship should not be allowed in public places. While most Sri Lankans,
including religious minorities, say they feel free to express religious opinions in their local area,
a noteworthy 13 percent of Muslims and 23 percent of Hindus do not feel free to express
themselves anywhere or only amongst friends and relatives. Muslims feeling restricted are
largely concentrated in Puttalam, Batticaloa, and Ampara.
o Government sponsored displays of religion are a source of tension. Significantly more
minorities than Buddhists believe government organizations should not display religious
symbols or signage. Sixty‐seven percent of Muslims and 66 percent of Hindus oppose
4
government use of religious symbols and signage compared to 46 percent of Buddhists. More
than twice as many Buddhists (38%) as Muslims (15%) believe government should be allowed
to display religious symbols or signage.
o A sizable minority of Muslims and Hindus do not reject the use of violence as a means to
achieve religious or political ends. Philosophical views on the use of violence are mixed and
vary by religious group. Approximately one out of every five Muslims (19%) believes that
violence is necessary to make a point or be heard. Roughly 1 in 3 Muslims and Hindus believe
violence is acceptable to settle political rivalries. Forty percent of Hindus and 34 percent of
Muslims believe violence is acceptable for punishing those whose actions offend their
religious values. Fifty‐three percent of Hindus and 35 percent of Muslims believe violence is
acceptable to uphold moral standards. While most Sri Lankans believe terrorism cannot solve
problems, a notable minority of Muslims and Buddhists do not outright agree with the notion
Twelve percent of the Buddhists and 6 percent of Muslims believe to at least some degree
that terrorism can solve problems.
METHODOLOGY
*See Annex II for full methodology report
The survey questions and structure were developed by The Asia Foundation and administered by Nielsen
Lanka Company (Pvt) Ltd between August 2 and September 23, 2011. The survey was administered across
nine provinces, and included 5553 face‐to‐face interviews yielding an overall margin of error of +/‐ 1.3% at
95% confidence level.
The sample is comprised of respondents from Muslim, Buddhist, Hindu and Roman Catholic1 communities
in Sri Lanka, and is designed primarily to provide an understanding of each of these four religious
communities at the national level. The sampling plan provides a wide spread across the country, covering all
provinces for Muslims, Buddhists, and Hindus. For Catholics, who constitute a relatively low percentage of
the Sri Lankan population, a group of provinces were targeted that contain over 80 percent of the total
Catholic population in the country. Additionally, to allow for analysis of some religious groups in selected
provinces, oversamples were done as follows: Muslims, Hindus, and Buddhists in the Eastern Province;
Muslims and Hindus in the Northern Province; and Muslims in the Western Province.
Overall, margins of error at the national level for the four religious groups are as follows: Muslims +/‐ 2.2%;
Buddhists +/‐2.7%; Hindus +/‐ 2.7%; Catholics +/‐ 3.5%. All samples were split 50%‐50% by gender.
1 Roman Catholics were selected to represent the Christian community as whole because they constitute approximately 90% of the Christian population.
5
Sample Size and Margin of Error by Religious Group
Population above
18 years Percentage
Sample Size Margin of
error %
Buddhist 8,592,000 69% 1344 2.7
Hindu 1,755,000 14% 1356 2.7
Muslim 1,185,000 10% 2070 2.2
Roman Catholic 898,000 7% 783 3.5
Total 4 groups 12,430,000 100% 5553 na
Respondents were randomly identified from citizens living in households (own/rent etc.) in all the provinces
during the survey time, aged between 18‐70, and belonging to the targeted religions—Buddhism, Hinduism,
Islam, and Roman Catholicism. All nine provinces were included in the survey however some parts in the
Northern Province were not included due to accessibility issues.
A stratified random sampling technique was utilized with the 5553 sample split by province and the
province sample further distributed across all districts. The sample in each district was again split in order to
cover the various Divisional Secretariats (DS divisions). In each DS division GN divisions were selected as
clusters using SPSS software random number generation process. For 5553 interviews, the total number of
sampling points was designated at 556, however 563 sampling points were taken to compensate for areas
in which interviews were not completed due to security related issues. Quotas were kept for each religious
group, and urban – rural split was maintained based on actual proportions.
In each selected sampling point, 10 interviews were targeted and broken into two sets by starting the
interviews from two separate starting points to enhance the representativeness of the sample. Five
interviews were conducted in each starting point in every GN until 10 interviews were completed. Popular
landmarks were selected as starting points for household identification and the interviewers used the right
hand rule of field movement while selecting the households. In urban areas every 3rd household was
selected while in rural areas every 2nd household was selected. Within households, respondents were
selected using a Kish Grid.
GPS handsets were used by field staff to mark the coordinates of the 563 sampling points across the
country where the clusters of 10 interviews were conducted. The GPS coordinates of the sampling points
serve to provide a reliable reading of the location and an indication of whether certain viewpoints are
geographically related. Maps demonstrating the geographic dimensions of opinions from the four religious
groups have been included in this report.
Overall, 22 percent of interviews were subject to quality control checks, and 10 percent of interviews were
accompanied by supervisors from the field partner organization.
NATIO
A sizab
minorit
feel th
directio
the rig
country
and jus
(21%) a
geogra
directio
Infrastr
the cou
right di
the righ
publiciz
by so
harassm
An exc
increas
wrong
ONAL MOOD
ble plurality of
ties and respo
e country is m
on. A quarter o
ht direction b
y is moving in t
st 22 percent o
and Hindus (2
phic trend als
on reside almo
ructure improv
untry is movin
irection or said
ht direction. Th
zed incidents i
called “grease
ment or issues
eption to this
ses in harassm
direction.
f Sri Lankans b
ondents in the
moving in the
of the country
ut others in t
the right direct
of Hindus say th
3%) as Buddhi
so emerges, s
ost exclusively i
vements and t
g in the right d
d some things r
he two most co
nvolving rando
ed devils”, rela
related to the
trend can be
ent/threats to
believes the co
near North an
e right directio
(26%) takes th
he wrong dire
tion. Forty‐nin
he country is m
ists (12%) beli
pecifically res
in Mannar, sou
the ending of t
direction. Resp
right/some thi
ommonly cited
om attacks—m
atively few re
e security of wo
seen in Musl
o security of w
6
ountry is movi
nd East are les
on, while just
he middle grou
ection. Fewer
e percent of B
moving in the r
eve the count
pondents who
uthern Vavuniy
the war are se
pondents who
ngs wrong wer
d reasons are in
mainly on wom
espondents cit
omen (6%).
im and Hindu
omen as a rea
ing in the righ
ss positive. Ov
15 percent fe
und, indicating
religious mino
uddhists comp
right direction.
try is moving i
o feel the cou
ya, Trincomale
een are the ma
indicated that
re asked why t
nfrastructure i
percent
war has
30 perce
Increase
increase
are the
feel the
wrong
respond
the pric
corrupt
reasons
going
Interest
survey
en—had taken
ted increases
communities,
ason for believ
ht direction, ho
erall, 43 perce
eel it is movin
g that some th
orities than Bu
pared to 38 pe
Nearly twice a
n the wrong d
untry is movin
ee, and Battica
ain reasons for
t the country is
they felt things
mprovements
t of responden
s ended which
ent of respond
es in the price
e in corrupti
e main reason
e country is
direction. Mo
dents cite eith
ce of goods (33
ion and bri
s they believe
in the wr
tingly, while a
was conducte
n place throug
in crime (6%)
, where about
ving the count
owever religio
ent of Sri Lanka
ng in the wro
hings are going
uddhists feel t
rcent of Musli
as many Musli
direction. A cle
ng in the wro
loa.
r those who fe
s moving in the
s were going in
s, cited by 35
ts, and that th
is mentioned b
dents.
of goods and
on and bribe
s for those w
moving in t
ore than half
her increases
3%) or increase
bery (20%)
e the country
rong directio
at the time t
ed several hig
hout the coun
) or increases
t one in six cit
ry is going in t
ous
ans
ong
g in
the
ms
ms
ear
ong
eel
e
n
e
by
an
ery
who
the
of
in
e in
as
is
on.
the
hly
try
in
ted
the
7
RELIGIOUS BELIEFS AND PRACTICES
Sri Lankans overwhelmingly perceive their society to be more religious today than five years ago.
Nationally, approximately two out of every three Sri Lankans believes that people are more religious today
than they were five years ago. Just 9 percent feel people are ‘a little’ or ‘much less’ religious than five years
ago. Broken down by religion, the data reveals that in all groups a majority perceive society to have become
more religious to at least some degree. By a significant margin, more Buddhists (70%) than other religious
groups indicate that people are “much more religious”. Muslims fall second behind Buddhists in perceived
increase in religiosity with over half
(53%) claiming to have become
much more religious. Hindus show
the lowest level of perceived
increase in religiosity with 39
percent saying they are much more
religious, while Catholics fall in the
middle with 44 percent much more
religious. Notable is that more
Hindus and Catholics also claim to
be less religious than other
religious groups, although the
proportion is still low at 13 percent
each (9% a little less religious, 4%
much less religious).
Most Sri Lankans prefer local religious leaders as their main source of guidance on religious matters with
the exception of Muslims who prefer holy books/Quran. Respondents were given the opportunity to
select multiple sources from which they obtain guidance on religious matters, and the results show that
more than three‐quarters (77%) consult their local religious leader, while 37 percent utilize holy books, and
22 percent other religious literature. Muslims deviate from other religious groups in that holy books are the
preferred source (76%). This
preference may be explained by
the high literacy rate in Sri Lanka
which allows for direct immediate
access as well as the reality that
that it is not customary for
Muslim women to interact with
Islamic leaders, making holy
books the default option for
essentially half of all Muslims.
Do you think people are more religious today than five years ago?
Most often sought source of guidance
Observ
indicat
cover p
Lanka c
women
and niq
the pa
percen
five yea
Female
respon
match,
choice
that Isl
with cu
becaus
vance of core
ors of religiosi
partially and p
cover, a 13 per
n cover, the su
qab (13%). Inte
ast five years
tage of wome
ars ago).
CURRENTLY W
e Muslim resp
dents were as
with the mos
(female 49% m
lam requires t
ulture” (female
se their husban
Observa
Muslim pract
ty are how ma
primarily as a m
rcent increase
urvey revealed
erestingly, the
has almost
n fully coverin
Whe
WEAR
ondents who
ked why the w
st common re
male 45%). Th
hat abaya/niq
e 7% male 8%)
nd or father ins
ance of core Isla
tices is very h
any times per d
matter of pers
compared to f
that most on
data also sho
entirely involv
g with abaya a
ether women ar
cover were a
women in their
sponse by bot
he second mo
ab must be w
). Interestingly
sists they do so
amic practices
8
high among Sr
day people pra
Near
simil
indic
sligh
claim
The
perc
indic
thos
unab
Thre
now
com
onal choice. C
five years ago.
ly wear abaya
ows that the in
ved adopting
and niqab stay
e/were wearing
sked why the
r house cover.
th genders tha
ost cited reaso
worn (female 2
, only 4% of w
o.
ri Lankan Mus
ay and how of
rly all Muslim
lar rate of
cate that they
htly lower bu
m to recite the
rate of perf
ent. While per
cator of religi
e who are e
ble.
ee quarters of
w cover repre
pared to five y
Currently, 75 pe
. While a large
(62%) rather
ncrease in cov
abaya only
ying essentially
g abaya/niqab?
FIVE YE
ey do so, and
The response
at covering is
n, again for bo
27%, male 30%
women indicate
slims. In Islam
ften they fast d
ms pray five
Muslim resp
fast during Ra
t still substa
e Holy Quran e
forming Haj o
rforming haj m
osity, it is no
either physical
f Muslim wom
esenting a ma
years ago, how
ercent of Mus
e majority of Sr
than fully cov
vering that has
(partially cove
y the same (11
EARS AGO
corresponding
es of males and
done as a ma
oth females an
%), followed by
ed that they w
m, two importa
during Ramada
times a day.
pondents (95
amadan , while
ntial proporti
everyday (72%
or Umrah is
may be seen as
ot obligatory f
lly or financia
men in Sri Lan
arginal increa
wever most on
lim women in
ri Lankan Musl
ering with aba
s occurred duri
ering), with t
% current vs. 9
gly male Musl
d females clos
atter of person
nd males, is th
y “in accordan
wear abaya/niq
ant
an.
A
5%)
e a
ion
%. ).
18
an
for
ally
nka
ase
nly
Sri
lim
aya
ing
the
9%
lim
ely
nal
hat
nce
qab
9
Most Muslims prefer to consult the Holy Quran to determine whether or not a certain action is against
the teachings of Islam. Respondents were asked if somebody were to tell them that a certain action was
against the teachings of Islam who or what they would consult for validation. Just over half (51%) of
Muslims would consult the Holy Quran, while roughly one quarter (26%) would consult their local imam.
The preference for consulting the Holy Quran versus imams or other Islamic leaders does not necessarily
reflect selective bias against local religious leaders for such matters. Rather preference might be largely
influenced by several factors relating to
accessibility. First, while in many
countries with Muslim populations rates
of literacy are low and the Holy Quran is
only available in Arabic, in Sri Lanka
literacy rates are high and Tamil language
versions of the Holy Quran are widely
available and therefore convenient
references. Second, women do not
customarily approach imams or other
types of Islamic leaders for advice,
leaving the Holy Quran or other holy
books as the default option for validating
teaching of Islam.
A majority of Sri Lankan Muslims believe
that at least to a certain extent traditions
and practices of Islam in Sri Lanka are
different from other countries with
Muslim populations. While Muslims don’t
feel strongly that their traditions and
practices of Islam are different than other
countries, they generally agree with the
notion, with 43 percent saying they
“somewhat agree” and 18 percent strongly
agree. Only 18 percent disagree that Sri
Lanka’s brand of Islam differs from other
countries.
If somebody were to tell you that a certain action was against the teachings of Islam, who or what would you consult for validation?
Agree or disagree: ‘Some traditions and practices of Islam in Sri Lanka are different from those of Muslims in other countries.’
10
Muslims generally view foreign influence
on their religious teachings positively,
while Buddhists view it negatively. A
plurality (38%) of Muslims view foreign
influence on their religious teachings
positively, while 18 percent say it is
neither good nor bad and 33 percent
view it negatively. This compares to just
9 percent of Buddhists who believe
foreign influence is a positive force on
their religion and a substantial 65
percent of Buddhists who feel it is bad
or very bad.
Muslims cite Saudi Arabia, India, and Pakistan as the countries having the greatest influence on their
religious values and practices. For Sri Lankan Muslims, Saudi Arabia is perceived as having the greatest
influence on their religious values and practices. The second most commonly cited country(s) of influence
for Muslims is Pakistan (18%) and India (18%). Meanwhile, 27 percent of Buddhists cite India, and 20
percent the United States. Regarding influence by the United States, given the absence of a commonly
shared religion with the vast
majority of Sri Lankans, it is likely
that respondents considered the
influence of the United States on
their religious values in the
context of general cultural
influence/international trends
(movies, television, etc), and
perhaps negatively so.
Is it a good thing for their religious teachings to be influenced by other countries or international trends?
Which countries have the greatest influence on religious values and practices in Sri Lanka?
11
ROLE AND INFLUENCE OF RELIGIOUS LEADERS
Muslims are relatively split in terms of who they consider to be their local religious leader. Just over half
of Muslims (52%) consider Maulvis to be their local leader while a nearly equal proportion (43%) identifies
Imams. Rather than indicating internal disagreement, this split in perception among Muslims likely reflects
the fact that in many communities Maulvis—who are primarily religious scholars—also often play the role
of Imam, leading prayers and providing spiritual
guidance. Among other religious groups,
Buddhists consider monks (99%) as their local
religious leader, Hindus point to priests (86%) and
Catholics to priests (67%) and bishops (27%).
Local religious leaders are by far the most
respected leaders in the country for Sri Lankans
of all faiths. Respondents were asked to gauge
the level of respect people have for a range of
local leaders in their GN division. A substantial 80
percent of respondents say their local religious
leaders are ‘very well respected’. This compares to 32 percent for GN officials, and just 12 percent who say
politicians are ‘very well
respected. Only 2 percent
say religious leaders are
‘little respected’ or ‘not
respected’ compared to
55 percent who say
politicians are little or not
respected. Perceptions
are relatively similar
across each of the 4
religious groups.
Respondents were asked to identify which leaders they believe are most active in taking steps to safeguard
the rights of minorities. Across all religious groups, religious leaders are unanimously (66%) viewed as being
most active, while local
and national politicians
are seen as highly inactive
in safeguarding religious
minority rights, cited by just
2 percent of respondents.
*Buddhists responding considering religious minorities. Others considering their own religious
community.
Who do you consider your local religious leader?
How well respected are the following leaders in your GN division?
Who ranks first among those taking steps to safeguard the rights of religious minorities?
12
A large majority of Sri Lankans believe that religious leaders should abstain from involving themselves in
politics, however nearly half of Muslims believe religious leaders should have some role in politics.
Overall, more than two‐thirds (66%)
of respondents believe that religious
leaders should have no role to play in
politics, while 20 percent say they
should have a small role and just 8
percent a large role. Muslims stand
out with more (42%) believing
religious should have at least some
role to play in politics (30% small role,
12% large role), while Buddhists are
the most averse to involvement of
religious leaders in politics (16% small
roll, 6% large roll).
While Sri Lankans strongly believe local
religious leaders should stay out of
politics, a staggering 84 percent think
that religious leaders should be involved
in the development of neighborhood
amenities. This unanimous view is
shared among Muslims, with 81 percent
saying religious leaders should have
either a major or at least a small role in
local development efforts.
Consistent with the belief that religious leaders should be involved in local development efforts,
respondents also strongly feel that
politicians should consult local
religious leaders in problem solving.
Eight in ten respondents agree (41%
strongly, 40% somewhat) that if
religious leaders are not consulted
on the problems facing an area,
political leaders will make mistakes.
Should religious leaders have a role in the development of neighborhood amenities?
‘If religious leaders are not consulted on the problems facing an area, political leaders will make mistakes’
Should religious leaders be involved in politics?
13
However, only a quarter say that religious leaders participate in local development efforts on a regular
basis, and the rate is even lower among Muslims. Despite overwhelmingly believing religious leaders
should be involved in community development efforts, just 26 percent say religious leaders in their GN are
actually participating on a regular
basis. This gap between desired
involvement and actual
involvement of religious leaders is
even greater among Muslims,
only 17 percent of which say their
leaders regularly (12%
‘frequently’, 5% ‘all the time’)
participate in local development
efforts.
Nearly three‐quarters of Muslims respondents or their family are currently attending or have attended a
religious school, and the rate is even higher among Buddhists. Respondents were asked whether or not
religious education is more
important today than 5 years
ago, and virtually all Sri
Lankans (89%) confirm that
religious education is more
important. This belief is
consistent across all groups,
including over 90 percent of
Muslims and Buddhists who
confirm increased importance.
The perception that religious education is more important today than five years ago is supported by actual
reported rates of attendance in schools providing religious education. Rates are highest among Buddhists,
with nearly nine in ten (87%) confirming current or past attendance. Attendance is also high among
Muslims, with nearly three quarters (73%) confirming, and only slightly lower among Catholics (70%). By a
significant margin, Hindus (42%) report the lowest rate of attendance.
Religious education among Sri Lankan Muslims is almost always pursued domestically, with only a miniscule
proportion (4%) indicating they have studied abroad, the vast majority having done so in Saudi Arabia. In
addition to asking Muslim respondents if they have travelled abroad to pursue religious education, the
survey also asked respondents to
specify which countries they
would hypothetically prefer to go
to for religious education Muslims
overwhelmingly point to Saudi
Arabia (62%), followed distantly
by India (5%), Pakistan (4%), and
Egypt (4%).
How often do religious leaders participate in development in your GN division?
Are you or members of your immediate family currently attending or have you/they attended a school providing religious education?
Which countries would you prefer to go to for religious education?
14
INTER-RELIGIOUS RELATIONS
A majority of respondents feel that the end of the war has not brought an end to ethnic conflict in the
country but that it has been reduced. Over half (53%) of Sri Lankans feel that the conclusion of the war has
not ended ethnic conflict in the country but that it has been reduced, while a third (32%) believe that the
conclusion of the war has indeed brought an end to ethnic conflict in Sri Lanka. Just 7 percent believe the
end of the war has not changed anything. Views are consistent across ethnic groups with the marked
exception of Hindus, among
whom fewer believe that the
conclusion of the war has
brought an end to the
conflict (21%) and more
believe that nothing has
changed (16%).
Sri Lankans believe that the
rights of minorities have
improved at least to some
degree compared to five
years ago, however Buddhists believe the improvement has been much more substantial than minorities
themselves. Just under half of all respondents (47%) feel the rights of minorities are much better now
compared to five years ago, while a third (32%) believes they are only somewhat better. Fifteen percent
feel rights are unchanged, while
just 4 percent feels they are
worse. Breaking the results down
by religious group, it can clearly
be seen that religious minorities,
especially Hindus and Muslims,
are markedly less positive than
Buddhists in evaluating trends in
minority rights. Thirty‐four
percent of Muslims feel the rights
of minorities have become much
better compared to 56 percent for
Buddhists, while a notable one in
four Muslims feel the situation has not improved or has become worse compared to just one in ten
Buddhists.
Has the end of the war in Sri Lanka brought an end to the ethnic conflict in the country?
How are the rights of minorities now compared to 5 years ago?
15
A majority of Sri Lankans believe people of all religions are treated equally in the government job market,
however a significantly larger proportion of Buddhists feel this way than minorities themselves. While
over two‐thirds of Muslims (66%)
believe that people of all religions
are treated equally in the
government job market, this
proportion is still markedly below
the 90 percent of Buddhist who
feel the same. Highlighting the
gap in perceived discrimination
between Buddhists and
minorities most starkly, only half
as many Hindus (45%) as
Buddhists (90%) believe people of
all religions are treated the same in the government job market, and roughly one in five Hindus believe they
face discrimination.
Religious minorities feel much more strongly than Buddhists that the government should make extra
efforts to protect their interests, including land and assets. While overall, a majority from all religious
groups including Buddhists support special government protections for minorities when it comes to land
and assets, the intensity of support varies considerably by group. Specifically, roughly twice as many Hindus
(50%), Muslims (41%), and Catholics (41%) than Buddhists (21%) strongly feel the government should take
extra steps to protect the
land and assets of
minorities. Furthermore,
markedly more Buddhists
disagree (29%) that
government should
provide special
protections compared to
minorities (Hindus 7%,
Muslims 11%, Catholics
16%).
Are people treated equally or do some face discrimination in the government job market?
‘Government should make extra efforts to protect the land and assets of religious minorities.’
RELG
While
ethnic
reserva
‘strong
lives fr
with re
to say
or ‘neit
Forty‐f
strongl
somew
percen
Virtual
with
roughly
(18%) d
Inter‐re
the cou
opinion
versus
the ge
accomp
Two lo
IOUS EXPRE
respondents f
and religious
ations among
gly agree’ with
eely. Less than
espondents mo
they ‘somew
ther agree nor
our percent o
y agree, 37
what agree,
t express
ly no Buddhist
the notion,
y one in fiv
disagree.
eligious/inter‐
untry. The sur
n, with a comb
38 percent wh
eographic loca
panying map r
cations of part
ESSION AND
from all religi
s groups and
religious min
the notion tha
n half of respo
ore inclined
what agree’
r disagree’.
of Muslims
7 percent
while 12
neutrality.
ts disagree
however
ve Hindus
‐ethnic tension
rvey sought to
bined 43 perce
ho disagree th
tion of respo
espondents wi
ticular note ar
TOLERANC
ous backgrou
d everybody l
norities than
at Sri Lanka is a
ondents from e
n related to re
o examine the
nt agreeing (17
hat it should n
ondents expre
ith strong opin
e Batticaloa an
‘Sri Lanka is a
16
E
nds generally
lives freely, t
Buddhists. Ov
a country for a
each of the thr
eligious expres
e issue of cons
7% strongly, 26
ot be allowed
essing a stron
nions are conce
nd Ampara, as
a country for all
believe that
the sentimen
ver two‐thirds
all ethnic and r
ree religious m
ssion is high i
struction of p
6% somewhat
(22% strongly
g opinion on
entrated in a n
s they possess
ethnic and relighere freely’
Sri Lanka is a
t comes with
of Buddhists
religious group
minority group
n specific geo
rominent relig
places
public ar
is increa
in
localities
Lanka, i
possibly
exclusivit
Responde
whether
disagree
statemen
statues
worship
allowed
public
Nationall
reveal
) that it should
y, 16% somew
the matter,
number of area
a concentratio
gious groups an
a country for
h notably mo
s (69%) say th
ps and everybo
ps strongly agre
ographic areas
gious symbols
of worship
reas given that
asingly comm
multi‐religio
across
in some cas
as a show
ty and streng
ents were ask
they agree
with t
nt “Religio
and places
should not
to come up
place
ly, the da
a near sp
d not be allowe
what). In terms
as seen in t
as in the count
on of Sri Lanka
nd everybody liv
all
ore
hey
ody
ee,
of
or
in
t it
mon
ous
Sri
ses
of
gth.
ked
or
the
ous
of
be
in
es”.
ata
plit
ed,
of
the
try.
ans
ves
both st
Annex
Interes
matter
religiou
are for
worshi
allowed
statues
Govern
Buddhi
percen
percen
govern
symbol
46
Additio
many
(15%)
be all
symbol
Severa
possess
that
organiz
locatio
govern
trongly for and
I for descriptio
stingly, the dat
exists. At the
us minorities, m
r it. In the cas
p should not b
d (42%) than n
s and places of
nment sponsor
ists believe go
t of Muslim
t of Hind
ment use
ls and signage
percent of
onally, more
Buddhists (38
believe gove
owed to dis
ls or signage.
al areas of
s concentratio
believe
zations should
ns where at
ment organiz
d against allow
on of these geo
a reveal that w
e same time, a
more responde
se of Muslims,
be allowed in
not (41%). Add
f worship than
red displays of
overnment org
ms and 66
dus oppose
of religious
e compared t
f Buddhists
than twice a
%) as Muslim
rnment shoul
splay religiou
the countr
ons of citizen
governmen
not display sy
least 60 perc
ations display
wing statues a
ographic locati
within each of t
a clear trend
ents are agains
, 45 percent a
public spaces
ditionally, more
minorities, esp
f religion are a
ganizations sho
to
s.
as
ms
d
us
ry
ns
nt
ymbols or sign
cent of respo
ying religious
‘Governme
17
and places of w
ons as well as
the four religio
emerges when
st allowing stat
agree while 30
. Conversely, a
e Buddhists ‘st
pecially Hindus
lso a source of
ould not displa
nage that is re
ndents either
signs/symbols
ent organization
worship to be
others mentio
ous groups, a s
n cross compa
tues or places
0 percent disa
among Buddh
rongly disagre
s (10% strongly
f tension. Sign
ay religious sy
eligious in nat
r strongly agr
s. This issue
ns should not dis religious in
e erected in pu
oned in this rep
similar lack of c
aring the four
of worship in p
gree that stat
ists more beli
ee’ (26%) with b
y disagree).
nificantly more
ymbols or sign
ure. The map
ree or strongl
can potential
indicator
between s
communities
Buddhist
organization
religious gro
they share
Geographic
showing
opposition
displaying r
or symbols
Batticaloa,
Vavuniya,
southern Nu
Jaffna.
display any symbn nature’
ublic places (S
port).
consensus on t
groups—amo
public areas th
tues or places
eve it should
banning religio
e minorities th
nage. Sixty‐sev
shows sampli
ly disagree w
ly serve as o
for tensi
some Buddh
s (and p
governme
ns) and minor
oups with who
public spac
locatio
the stronge
to governme
religious signa
include Ampa
southe
Mannar islan
uwara Ellya, a
bols or signage
See
the
ong
han
of
be
ous
han
ven
ing
with
one
ion
hist
ro‐
ent
rity
om
es.
ons
est
ent
age
ra,
ern
nd,
and
that
Inter‐re
publiciz
current
with th
(64%) o
tolerat
consist
groups
slightly
18 per
agree
conver
tolerat
agree,
agree).
while
conver
throug
the
indicat
regardi
of relig
reveal
this tim
conver
Overall
area, a
eligious tensio
zed in recent y
t feelings abou
he statement,
of Sri Lankans
ed (47% stro
ent across rel
, with only Cat
y less adamant
cent of Sri La
that rel
sion should
ed (9% str
9% some
Geograph
opposition
sion is wides
hout the co
same loc
ing te
ing public di
gion again app
religious te
me on the iss
sion.
l, while most S
noteworthy m
on and confli
years, particul
ut conversion,
“Converting p
disagree with
ngly disagree,
igious
tholics
t. Just
nkans
igious
be
rongly
ewhat
hically,
to
spread
untry,
ations
ension
splays
ear to
nsion,
sue of
ri Lankans (91%
minority of Mu
ct resulting f
arly in the aft
and according
people from o
h the idea that
, 17% somew
%) feel free to
uslims and Hin
18
from attempts
ermath of the
ly respondents
one religion to
t attempts to
what disagree)
express religio
dus feel restri
s at religious
e 2004 Tsunam
s were asked w
o another sho
convert peopl
). This anti‐co
ous opinions a
cted. About o
conversion h
mi. The survey
whether they a
ould be tolera
le from one re
onversion fee
lmost everywh
ne in eight Mu
one in
not fee
themse
or only
and re
feeling
largely
Puttalam
and Am
has been hig
sought to gau
agree or disagr
ted”. Two‐thir
eligion should
ling is relativ
here in their lo
uslims (13%) a
four Hindus
l free to expre
lves anywhe
amongst frien
latives. Musli
restricted a
concentrated
m, Batticalo
mpara.
hly
uge
ree
rds
be
ely
cal
and
do
ess
ere
nds
ms
are
in
oa,
VIOLE
When
either
of Hind
Even a
violenc
Geogra
violenc
suggest
Battica
exclusiv
primar
inspire
cited
concen
respon
areas
southe
and Am
questio
distinct
ethnica
inspire
results
ethnici
often
interch
Lanka.
ENCE AND EX
it comes to vi
criminal activ
dus, while relig
among Catholi
ce. Notably few
aphically, the
ce are concent
ted in maps pr
loa, Akurana,
vely in areas w
ily ethni
d violence th
by l
ntrations
dents. T
include Man
rn Nuwara
mpara. While
on inferred
tion betw
ally and religio
d violence,
demonstrate
ty and religion
discu
hangeably in
What types o
XTREMISM
iolence in the
vities (59%) or
giously inspire
cs, a combine
wer Buddhists c
accompanying
rated in many
resented in th
Trincomalee,
where respond
cally
at is
arge
of
hese
nnar,
Eliya
the
a
ween
ously
the
that
n are
ssed
Sri
(Note: be
offset to e
of violence are t
country, while
r political ope
d violence is c
ed 22 percent
cite these two
g map reveals
of the same lo
e previous sec
, and Kalutara
dents also poin
cause of locatio
ensure all results
taking place in t
19
e Sri Lankans
erations (25%)
cited by 12 pe
t point to rel
types of violen
that respond
ocations where
ction. Areas wh
a. While relig
nted to ethnica
n overlap for th
s are visible)
the country?
perceive the
), a notable 1
ethn
freq
type
viole
Mus
roug
Brok
viole
perc
ercent of Musli
ligiously (11%)
nce (ethnically
ents citing re
e tension relat
here both type
giously inspire
ally inspired v
hese two types o
majority of it
15 percent ci
nically inspired
uency of repo
es of likel
ence is even
slims and H
ghly a quarte
ken down, et
ence is men
cent of Muslim
ims and 10 pe
) or ethnically
y 7%, religiously
ligiously or et
ting to religiou
es of violence
ed violence w
violence, in som
of violence, dots
to be related
te religiously
d violence. T
orting these tw
ly interrelat
higher amo
Hindus, totali
er of respons
hnically inspir
ntioned by
ms and 15 perce
ercent of Hind
y inspired (11
y 4%).
thnically inspir
s expression w
are cited inclu
was cited almo
me locations it
have been sligh
to
or
The
wo
ted
ong
ing
es.
red
14
ent
us.
1%)
red
was
ude
ost
t is
htly
Despite
Sri Lan
the cou
extrem
violenc
many
where
implied
section
Island,
Ampara
Akuran
descrip
geogra
Muslim
Muslim
‘Sr
‘The
e perceiving m
nkans indicatin
untry is vulner
mist views
ce are concent
of the same
religious tens
d in the p
n, including
southern Va
a, Batticaloa
na. (See Anne
ption of
phic locations)
ms are divided
ms outright ag
ri Lanka is vuln
ere are differencmatters of rel
most current ac
ng that
rable to
and
rated in
e areas
ion was
previous
Mannar
avuniya,
a, and
ex I for
these
)
regarding the
ree (2% stron
nerable to extrem
ces of opinion aligion that are le
ts of violence
e existence of
ngly, 10% som
mist religious vi
among some Mueading to confli
20
as either crimi
intra‐religious
mewhat) with t
iews and violen
uslims on ict.’
inal or politica
Sri La
is vu
from
Amo
belie
while
vulne
Hind
agre
resp
aver
vulne
s conflict in th
the assertion
opinion am
religion tha
interestingly
to express
assertion (
somewhat
percent are
answer (
geographic
percent of M
a large prop
Puttalam.
nce’
l in nature, a m
ankans feel th
lnerable to vio
m extremist
ong Muslims
eve the count
e just 21 perc
erable. Both
dus expre
ement (25%
ectively) than
age (20%) tha
erable.
heir communit
that there ar
mong Muslims
at are leading
y, less than ha
s disagreem
26% strongly
disagree), an
e neutral (21%
(22%). One
finding is tha
Muslims expres
portion resides
majority (51%)
hat their coun
olence stemmi
religious view
, 53 perce
try is vulnerab
cent say it is n
Muslims a
ess strong
% and 29
n the nation
at the country
ty. Relatively fe
re differences
on matters
to conflict. Y
alf (45%) choo
ent with t
disagree, 19
d a further
%) or do not
e notewort
t among the
ssing agreeme
s in Ampara a
of
try
ing
ws.
ent
ble
not
and
ger
9%
nal
y is
ew
of
of
Yet
ose
the
9%
43
to
thy
12
nt,
and
Geogra
expres
Hindus
uphold
punish
religiou
politica
eschew
sizable
and Hin
not re
violenc
circum
one in
(2%
somew
someti
use vi
point o
percen
strongl
feel th
accomp
concen
aphic areas of
s support for
s. Philosophica
moral stan
those off
us values, or to
al rivalries is
wed, howev
minority of M
ndus in particu
eject the u
ce in
stances. R
five Muslims
strongly,
what) that
mes necessa
olence to m
or be heard, w
t of Hindus
y, 12% some
he same. This
panying map
ntrated in Man
the country c
the use of vi
l views on the
ndards,
fending
o settle
largely
er a
Muslims
ular do
se of
these
Roughly
agree
17%
it is
ary to
make a
hile 15
s (3%
ewhat)
s compares to
shows, Musli
nar Island, Sou
consistently su
iolence for re
use of violenc
o just 5 perc
ms who feel
uthern Vavuniy
21
uggesting high
ligious or pol
ce are mixed an
ent of Buddh
violence is s
ya, Ampara, Tr
levels of relig
itical ends, es
nd vary by relig
hists (2% stro
sometimes nec
incomalee, an
gious tension a
specially amo
gious group. U
ongly, 3% som
cessary to ma
d Puttalam.
Furthermor
Muslims (3
(33%) belie
acceptable t
rivalries. Ju
Buddhists
Muslims ex
for the use
settle politi
largely co
Mannar,
Vavuniya,
Ampara, a
Hindus exp
are con
Batticaloa,
southern N
Jaffna.
are also likely
ng Muslims a
Use of violence
mewhat). As t
ake a point a
e, roughly 1 in
1%) and Hind
eve violence
to settle politi
st 8 percent
feel the sam
pressing suppo
e of violence
ical rivalries a
oncentrated
Southe
Trincomale
and Batticalo
pressing suppo
ncentrated
Badu
uwara Ellya, a
to
and
e to
the
are
n 3
dus
is
cal
of
me.
ort
to
are
in
ern
ee,
oa.
ort
in
lla,
and
Forty p
actions
Manna
Vavuni
Trincom
and A
express
violenc
those
offend
values
concen
same a
that in
for vio
politica
Badulla
southe
Ellya, a
Fifty‐th
standa
percent of Hind
s offend their r
r, Southe
ya,
malee, Ampar
Akurana. Hind
sing support f
ce to puni
whose actio
their religio
a
ntrated in th
areas as Hind
dicated suppo
olence to sett
al rivalrie
a, Batticalo
rn Nuwa
nd Jaffna.
hree percent o
rds. Muslims
dus and 34 per
religious value
rn
ra,
us
for
sh
ns
us
re
he
us
ort
tle
es:
oa,
ara
of Hindus and
who indicate
cent of Muslim
es. Muslims wh
35 percent of
that violence
22
ms believe viole
ho indicate tha
f Muslims beli
e is acceptab
ence is accepta
at violence is a
ieve violence
ble are conce
able for punish
acceptable are
is acceptable
ntrated in M
Vavuniya,
Ampara,
Hindus sup
of violen
concentrat
four areas
previous
regarding
violence:
Badulla, so
Ellya, Badu
hing those who
e concentrated
to uphold mo
annar, Southe
Trincomale
and Akuran
pporting the u
nce are aga
ted in the sam
identified in t
three ma
support f
Batticalo
outhern Nuwa
ulla, and Jaffna
ose
d in
oral
ern
ee,
na.
use
ain
me
the
aps
for
oa,
ara
a.
While
Buddhi
statem
that te
terroris
Meanw
and a f
most Sri Lank
ists do not agr
ent, “Terrorism
errorism canno
sm cannot sol
while, an even
further 4% neit
kans believe t
ree with the n
m cannot solve
ot solve proble
lve problems,
greater propo
ther agree nor
terrorism cann
otion. Respon
e problems”. T
ems. However
and a furthe
ortion (12%) of
r disagree. Wh
23
not solve prob
dents were as
The data revea
r, a notable 6
r 6 percent n
f Buddhists dis
ile Muslims re
blems, a nota
ked whether t
al that the vast
percent of M
neither agree
sagree that ter
espondents exp
able minority
they agree or d
t majority of S
Muslims say th
nor disagree w
rrorism cannot
pressing suppo
terrorism
problems
few in
degree
concentra
them can
namely
Kandy‐Aku
Kalutara,
Trincomal
who belie
solve p
concentra
Gampaha,
Kalutara.
of Muslims a
disagree with t
Sri Lankans agr
ey disagree th
with the notio
t solve proble
ort for the use
to so
are relativ
number, som
of geograp
ation amo
n be discerne
in Mann
urana are
a
ee. Buddhi
eve terrorism c
problems a
ated
, Kegalle, a
and
the
ree
hat
on.
ms
of
lve
ely
me
hic
ong
ed,
ar,
ea,
and
sts
can
are
in
and
24
CONCLUSIONS AND IMPLICATIONS
Ten geographic areas of Sri Lanka were identified in which findings from the survey suggest the
existence of ethno‐religious tension and potential for localized violence. Going forward, it will be
important to triangulate the findings for each of these “hot spots” with secondary sources and to
conduct in‐depth qualitative research to better understand key issues and concerns emerging from
these areas. One step has already been taken through a qualitative study following the survey (see
Annex 1) in 6 of the 10 hot spots identified, which examined unique factors in each location that
may be fuelling conflict as well as common structural factors among all locations such as level of
diversity and nature of segregation. Further research will also be necessary to determine the
political economy of conflict in the ten hotspots, more specifically whether or not there are actors
at the local level that are engineering tension and conflict for political gain. Finally, follow up
surveys will also be required to track indicators of tension and violence in the hotspots, as well as
to identify new areas that may be emerging.
The survey also highlighted that religious education is very important for Sri Lankans of all faiths,
and that the vast majority are currently attending or have attended a school providing religious
education. Religious institutions play a critical role in delivering religious education. Going forward
the Foundation would like to work with the Ministry of Religious and Cultural Affairs (MRCA),
specifically the Department of Buddhist Religious & Cultural Affairs and the Department of Muslim
Cultural Affairs, to develop a curriculum on religious tolerance with the aim of institutionalizing the
curriculum as part of the regular training provided through religious institutions. This will involve
the development of two curriculums: one for Buddhists and one for Muslims. The curriculum
development process will be led by two national‐level religious institutions, representing the two
religious communities, that will work closely with the MRCA and religious and social leaders from
the 10 identified locations (‘hotspots’). In addition, religious education is also being delivered
through the formal public education system where it has been thoroughly integrated with secular
curriculum and daily activities. Public schools are now largely segregated based on religious
affiliation, and the central role that religion plays in formal education is likely a key driver of
increasing religiosity and formation of more narrow or insular ethno‐religious identities.
Accordingly, steps should be taken to introduce curriculum into the public school system that
promotes religious tolerance.
Lastly, the survey revealed that religious leaders are highly influential among all religious groups
and there is a strong desire among the public for greater involvement of these leaders in improving
local social and economic conditions. The desire for greater involvement is likely driven by the
critical but ad hoc role the public has witnessed religious leaders playing in promoting
reconciliation and mediating of new conflicts between ethno‐religious communities both during
the war and in the current peace‐building process. Because religious leaders have been a
moderating influence on deep ethno‐religious divisions that characterize many localities in the
country, going forward it will be important that support be given to them to jointly work with local
officials in such critical interventions on a more systematic level.
25
ANNEX
I Qualitative Study on Inter‐Faith / Inter‐Ethnic relations in selected communities in
Sri Lanka
1. Introduction
The Vulnerable Population Survey conducted by the Asia Foundation in 2011 pointed to significant inter‐
religious, inter‐ethnic tensions in a number of geographical locations, namely Ampara, Baticaloa, Vauniya,
Mannar, Anuradhapure South, Akurana and Kaluthara‐Beruwala. The purpose of the present in‐depth study
was to conduct a series of interviews with a cross selection of people in these areas to determine the
underlying factors contributing to apparent tension in the above areas. More specifically attention was
focused on a number of questions, some exclusively focused on the Muslim community.
The survey questions were as follows;
a.) Religiosity: why people from across the four religious communities are more religious today
b.) Religious guidance: Why Muslims rely on holy books for guidance
c.) Practice of Islam: Why Muslims believe that the tradition and practice of Islam in Sri Lanka is different
from other countries with Muslim populations
d.) Religious leaders: Why religious leaders are more respected than other leaders (political leaders,
professionals, Grama Niladhari Officers etc) in society
e.) Violence: what people mean by violence; what kind of violent incidents are occurring; why people
believe that violence is necessary to make a point or be heard; and why violence is accepted to punish
those whose actions offend their religious values.
f.) Extremist views: why Sri Lanka is vulnerable to extremist religious views and violence
g.) Intra‐religious conflict: Specifically amongst Muslims
Interviews were conducted with a view to finding answers to these seven questions, besides finding out
what sort of inter‐religious, inter‐ethnic relations are prevalent in the respective areas and what sort of
local circumstances shape such relations.
This report presents the results of the qualitative field research conducted over the last few weeks in the
selected areas. The report is divided in to four sections. In section one, a brief account of the methodology
of the study is presented. Section two provides socio‐economic and demographic sketches of the field sites.
In section three results of the in‐depth interviews are analyzed. And finally, section four examines the
circumstances that led to inter‐community tensions in the areas covered by the qualitative study.
2. Methodology:
26
Field work was limited to a small number of in‐depth interviews in each selected locations. However in
order to ensure the quality of the data collected, three experienced research assistants were deployed in
the field under the guidance of a senior researchers. The three research assistants were drawn from the
three main ethno‐religious groups in the country, namely Sinhalese‐Buddhist, Hindu‐Tamil, Islamic‐Moor.
The senior researcher who accompanied the research assistants in the field was trilingual and speaks
Sinhala, Tamil and English. Given the short duration of the field work, it was felt that ethno‐religious
background of the research team would help establish rapport with the respondents more easily.
The key research questions were discussed by the research team before their commencement of the field
work with a view to formulating some guidelines for interviews. Initial field work was conducted at
Kaluthara and de‐briefing session was organized the following day to review the outcome of interviews.
This helped clarity issues and prepare the research team for subsequent interviews in other field sites.
The respondents for interviews were selected in such a way as to ensure this diversity in terms of age,
gender, class, ethnicity, religion, education and occupation. Interviews were conducted in the language that
the respondents found themselves to be more comfortable with the members of the research team were
able to conduct their interviews in any of the three languages spoken in Sri Lanka. Interviews were written
up as detailed interview reports by the research assistants following the interviews based on notes based
on notes taken down during the interviews.
A total of sixty one interviews were conducted at seven field locations. Several de‐briefing sessions were
conducted following the completion of interviews. The data collected during fieldwork was discussed at
length at these de‐briefing sessions in order o draw conclusions from the data.
3. Socio – economic and Demographical Background In this session, we provide brief sketches of the several field sites covered by the present study. The sites
were drawn from the North, East, Southern, and Central Provinces of the country. The distributions of field
locations are given below.
Province Locality
Northern Province Vavuniya , Mannar
Eastern Province Ampara , Baticaloa
Central Akurana
Southern Kaluthara, Beruwala
Vavuniya is the southern gateway to the Northern Province. It’s population is predominantly Tamil, though
a Sinhala minority is also found there. There is a small Muslim minority in the Vavuniya town which is a
growing urban centre in the region.
Mannar is largely inhabited by Tamils but a majority of them are Roman Catholic by faith. There is a sizable
minority of Muslims and the rest of the population is mostly Hindu Tamils. A majority of residents in
Mannar island depend on fishing and related activities, while those resident in the mainland areas in the
district are mostly engaged in agriculture. Muslim residents in Mannar had to caused the area during the
27
war due to the threat posed by the LTTE. Following the end of the war, many Muslims have returned to
their old resettlement.
Ampara Town is almost entirely inhabited by Sri Lankan Buddhists, while there are also Tamil‐Hindu as well
as Muslim villages in the surrounding areas. Some of the other towns in the district of Ampara are
dominated by Muslims. These include Kathankudi, Kalmunai, Saindamarthu, Samanthurai and Akkaraipattu.
Majority of the inhabitants in ampara district (42%) are Muslims while the remainders are almost equally
divided between the Sinhalese and the Tamils.
Baticaloa town is inhabited mostly by Tamils though there is a minority of Muslims both in urban as well as
rural areas. With the improvement of roads connecting Ampara with other towns, transportation of goods
and people between the two towns has increased tremendously, leading to greater interaction between
communities.
Akurana is a predominantly Muslim town in the Central province, surrounded by mostly Sinhala‐Buddhist
villages. The town is a thriving business centre and most of the businesses are owned by Muslims.
Sinhalese‐Buddhist villagers living in the surrounding areas are engaged in agriculture and depend on the
town for various day today needs.
Kaluthara – Beruwala area is inhabited by both Sinhala‐Buddhists as well as Muslims. Yet they live in almost
exclusive ethno‐religious neighbourhoods. Muslims are mostly concentrated in urban centres, while the
Sinhalese‐Buddhists are largely settled in the surrounding villages. Kalutara town is a well known Buddhist
centre due to the presence of a landmark Buddhist monument that attracts Buddhist worshippers from the
surrounding areas as well as from other parts of the country. On the other hand, Beruwala is a well‐known
Islamic Centre.
As mentioned earlier, the present study has been focused on seven specific questions that have arisen from
the Vulnerable Populations Survey. While some of these four questions have been common to all religious
groups in the country, the remaining three being related to the Muslim community. In the next section, the
discussion will initially focus on the four common questions. The section will conclude with several general
conditions based on the data.
4.1 Increased Religiosity among Sri Lankans
In depth interviews conducted among all major religious groups, namely Buddhists, Hindus, Muslims and
Christians, indicate that there has been an increase in religiosity among all of them. This is understandable
in view of the fact people belonging to all religions have been exposed to similar circumstances in the
recent past.
One of the key factors that has contributed to the above state of affairs is an increased sense of insecurity
and uncertainty created by the war and the major natural disasters like the 2004 Indian Ocean Tsunami.
While it is true that only certain section of the country is population was directly affected by the war and
the Tsunamis, the scale of the conflict and the Tsunami was such that almost the entire population of the
country was affected in one way or another. The mass media, in particular the electronic media played a
major role in instilling a deep sense of fear and insecurity among the people when graphic and horrific
28
details of the consequences of the war and the disaster were repeatedly displayed on TV screens on a daily
basis.
The population in the North‐East, directly exposed to the war, naturally lived in constant fear and
uncertainty for an extended period of time. Families, relations and friends of security forces personnel
stationed in the war zone also lived in constant fear. The deaths, disappearances and injuries of combatants
continued to be a regular occurrence during the period of the war while individual families conducted
religious rituals in the villages to bestow divine blessings on their loved ones, mass public rituals organized
by religious organizations, often with state patronage, were also common in many part of the country. Such
heightened religious activity contributed to increased religiosity among people.
In more structural terms, the social transformation that followed economic liberalization in the late 1970’s
appears to have been another major factor contributing to increased religiosity in the country. Growing
economic pressure, increasing mobility of people between rural and urban areas and mass migration of Sri
Lankan workers for employment overseas often leading to family disorganization and greater uncertainty
have affected many people in the country irrespective of their ethno‐religious background. Many religious
leaders have focused attention on the issues arising out of the above social transformation. The emergence
of Hela Urumaya led by several Buddhist monks, in particular the late Gangodawila Soma Thero, was partly
facilitated by the ‘moral outrage’ given rise to by the adverse consequences of rapid social and economic
change that followed the introduction of open economic policies. Religious leaders belonging to all major
religions expressed their concern in their religious sermons. This was clearly evident in Buddhist sermons,
particularly on television.
It is also significant that the electronic media became a major instrument for propagating religious functions
and talks. This has been particularly so in the case of Buddhism. Leading Buddhist monks appear on
television regularly for both sermons as well as public discussions on social issues. Major religious events
are telecast live on TV, reaching out to the wider population instantly. What is also noteworthy is that a
number of Buddhist monks emerged as leading figures attracting a mass following for their sermons and
public rituals, largely with the support of media institutions.
The North East war also increased the visibility of religious leaders both at a local and national level.
Through various initiatives such as inter‐faith dialogue on reconciliation and co‐existence, religious leaders
belonging to different religious groups began to play a significant mediatory role at a community level
where the conflict had led to inter‐community violence. Such efforts had brought people affected by the
war closer to their respective religions.
Increasing religiosity among people belonging to all religions has made distinct religious communities more
visible particularly where diverse religious groups live side by side in the same geographical area. In fact
increasing religiosity has made their differences starker, often leading to inter‐religious competition. The
construction of prominent religious symbols in public places, use of loudspeakers to reach out to a larger
audience when religious rituals or sermons are conducted, etc are quite common in multi‐religious
localities. The resulting greater sense of competition in turn appears to reinforce religiosity among people.
4.2 Increasing Respect for Religious Leaders
It is in the context of increased religiosity that the evident respect for religious leaders needs to be
examined. While it is true that there has been a certain erosion of credibility of some religious leaders at a
community level due to instances of corruption and immoral behaviour on the part of local religious
29
leaders, the close attachment to religion has enhanced their respect for religious leadership in general. This
is particularly so among Buddhists. The general disillusionment with political leaders and alienation from
public institutions such as the judiciary, state bureaucracy and law enforcement agencies have contributed
to enhanced respect for religious leaders. The fact that many religious leaders have played a mediatory role
in conflict areas by working together across religious divisions has also earned them considerable respect
from the people. Many religious institutions have provided shelter to civilians fleeing from violent conflict.
As mentioned earlier, religious leaders have paid greater attention to existential problems of the ordinary
people. They have emphasized the need to follow religious prescriptions to avoid trouble and bring up
children in a religious environment. State patronage extended to religious institutions to propagate
religious education of children has also enhanced the influence of religious leaders in society. While
children’s attendance in religious classes on Sundays has become the norm, in all part of the country,
exposure to religious teaching has also persuaded younger generations to respect religious leaders more
than other leaders.
4.3 Greater Tolerance of Violence as a means of Attaining Objectives
The respondents from different religious persuasions agree that their religions do not condone violence.
Yet, they see increasing competition among religious groups for public goods such as land and public space
leading to use of force, at times resulting in physical violence. This is particularly so in areas where ethno‐
religious groups live in segregated communities. In fact, this pattern is prevalent in many of the field areas
surveyed. In a few localities, religious groups live in mixed neighbourhoods where closer inter‐community
relations have evolved and people belonging to different religious groups do not perceive each other as
competing or conflicting groups. Inter‐group marriages and sharing of public goods such as schools have
facilitated inter group co‐operation. This was observed in Vavuniya, particularly among Buddhists and
Hindus. Another locality where such inter community cooperation has evolved is Akkaraipattu in Ampara
District.
The use of violence to achieve political objectives has been common in many parts of the country in the
recent past. JVP violence in the South and LTTE violence in the North and East as well as elsewhere in the
country, use of force by para military groups to deal with political decent are too well known to need any
elaboration. On the other hand, deprived or agitated communities and groups have also tended to use
aggressive means to send a message to authorities. The cases in point are protests against law enforcement
officers who fail to act impartially or violate the law, agitations against public authorities that failed to
maintain public amenities such as roads, violent demonstrations stated by unemployed graduates seeking
government jobs and workers protesting against perceived injustices.
Instances of violence mentioned above usually do not involve ethno‐religious groups as rival parties. On the
other hand, violent incidents such as physical assault, arson, etc that take place following road accidents
often turn in to inter‐community rivalries depending on the location of the accident and the parties
involved. For instance, if a vehicle driven by a member of one community is involved in an accident leading
to an injury or death of a person belonging to another community, the members of the latter community
that come forward to defined the interest of the victim, may react violently against the offending driver.
Respondents referred to many such incidents in Kalutara, Ampara and Akurana. Such incidents indicate the
prevailing, underlying tension between ethno‐religious communities in the areas concerned.
30
The underlying inter‐community tensions do not always lead to violent incidents. The likelihood of violence
depends on the prevailing power balance between communities, the nature of community leadership and
the spatial distribution of the respective communities.
In Kalutara/Beruwala area Muslims are concentrated in urban centres while the surrounding villages are
inhabited predominantly by Sinhala‐Buddhists. In the past, several violent incidents have occurred after
traffic accidents in urban centres when Sinhala‐Buddhist drivers of the vehicles were assaulted by Muslim
groups. Following such accidents, retaliatory attacks have occurred when Muslim persons drive through
Sinhala‐Buddhist dominated areas. Given the apparent balance of forces in the area, there has evolved a
tacit understanding among community leaders that it is best to avoid confrontation. However, the situation
is different in areas like Ampara and Akurana, where the level of tension between Sinhala‐Buddhist and
Muslims is high. While Ampara town is dominated by Sinhala‐Buddhists, several thriving towns in vicinity
like Kalmunai, Kalavanchikudi and Saindamarusu are mostly inhabited by Muslims. Sinhalese businessmen
and transporters have to drive through these Muslim towns to reach Baticaloa for business and other
purposes. If they are involved in any accidents hurting local residents in their towns, they face the threat of
violent attacks by local Muslim groups. On the other hand, many respondents point out that the Sinhalese
Buddhist elites control Ampara town and prevent Muslim businessmen from getting a foothold there.
Akurana town in the central province is dominated by Muslims, while many villages around the town are
inhabited by Sinhala‐Buddhist peasants who depend on the town for marketing their produce and buying
household goods. Many villagers feel powerless against Muslim domination in the town.
Mannar is quite different from other field areas due to the fact that the Tamil community there is divided in
to Hindu and Catholic communities. Muslims also constitute a significant minority there. There is
considerable tension between religious groups there. Muslims who fled the area due political violence have
been returning to their former habitats following the end of the conflict. They often receive special
assistance due to the intervention of their political leaders. This has apparently generated some tension
between them and their Tamil neighbours who are also struggling to emerge following the end of the war.
Some of the Tamils are refugees themselves returning from India and internal refugee camps.
There is also inter‐religious conflict between Catholics and Hindus. Some poor Hindus are allegedly
converted to Christianity and the Hindus appear to resist this trend by denying legitimate land rights to the
converted. They want the converted families to leave the area.
Though there is considerable tension between religious groups in Mannar, people seem to desist from the
use of violence to demonstrate their frustrations. This is due to a general tendency among all groups to
avoid conflict against a backdrop of violent conflict in the recent past that disrupted their lives in an
unperfected manner. On the other hand, increasing intergroup competition leading to the marginalization
of some has created a sense of injustice and discrimination in the minds of people who are adversely
affected. This is evident in many areas where interviews were conducted, particularly in Mannar, Baticaloa,
Ampara and Akurana.
The strong attachments to religion and religious practises have persuaded people to defend the religious
institutions and their right to engage in religious activities undeterred by outsiders. Attacks on religious
institutions, priest and devotees in the past have reinforced the above tendency. Many people expressed
the opinion that use of force is legitimate to defend their religion, religious institutions and practices. A high
level of religiosity among the member of one religious group tends to reinforce a similar tendency among
31
the members of other religious groups. This is particularly evident in Kalutara, Akurana, Mannar and
Ampara.
4.4 Extremist Religious Views;
There is no consensus among respondents regarding the meaning of extremist religious views. While some
consider strong attachment to religious doctrine as a form of extremism others refer to fundamentalist
groups either within or outside the major religious traditions. This division of opinion could be observed
among most respondents irrespective of their religious persuasions.
Strict adherence to Islamic teaching and associated social and cultural practices is the norm that most
Muslim respondents do not want to deviate from. Some of the deviations are due to the differences in
religious beliefs and practices among Muslims themselves. Others are due to the need for adapting to local
conditions. For instance, Muslims in some locations support Buddhist temples and work closely with
Buddhist devotees without moving away from their own religious beliefs and practices. Such instances were
evident in Baticaloa, Anuradhapura and Vavuniya. Such adoptions reduce the gap between religious groups
and facilitate co‐existence.
On the other hand, in the areas where population is highly segregated on religious lines and distinct
religious groups assert themselves in their public display of exclusivity and strength, people feel that there
is some form of religious extremism there.
Though there have been certain deviations from the mainstream religious traditions, these deviations are
not necessarily regarded by people as extremist tendencies. People feel that mainstream traditions are
strong enough to remain unchallenged by such ‘deviant’ groups and practices.
4.5 Intra‐ Religious Divisions
While there are divisions within all major religious traditions, they are not clearly visible or important at a
community level. In many cases, different religious practices within the tradition take place at a local level
and the devotees do not perceive it as an issue. Only one major conflict arising out of intra – religious
division among Muslims was recorded and this was at Kalutara where two Muslim factions clashed over the
issue. Among Buddhists and Hindus, diversity of religious practices is quite common and the devotees do
not perceive such diversity as a major issue.
Intra – religious diversity among Muslims are partly the result of external influences and partly due to
internal differentiation. The following distinct religious grouping were evident in the localities surveyed.
a.) Thableek Jammath b.) Thowleeth Jammath c.) Jama athe Islami d.) Thareeka e.) Tharkiya f.) Sunnathul val Jammath g.) Thaaha
32
Thableek Jamath is the largest group and is dominant in many areas, particularly in rural areas. Those who
follow this largely oral tradition are more conservative and are not very much exposed to modernizing
forces. On the other hand, Thowleeth Jammath is more urban and is adhered to by more educated,
professionally oriented Muslims. They are exposed to the outside world and rely on textual material for
guidance. Other groups have a small following and are not found every where.
Jama athe Islami and Thareeka are the result of Indian influence and display such influence in the way they
conduct their prayers.
Some of these divisions have led to violence, particularly between Thableek Jammath and Thowleeth
Jammath.
Tharkiya group is found only in Kaluthara area. They believe that dead people act as mediators between
People and God. In 2001, there had been a major clash between members of these groups and Thowleeth
Jammath members leading to two deaths and several injuries.
Besides the main Muslim groupings namely Thableek Jammath and Thouheeth Jammath,there was a small
group identified as Thaaha in Batticoloa whose members follow the fasting schedule derived from Saudi
Arabia.
In spite of the presence of small Muslim religious sects in several parts of the country, the main division
appears to be between more traditional Thableek Jammath and civil society oriented Thowleeth Jammath.
These internal divisions are important within the Muslim community but do not appear to undermine their
religious and ethnic identity which is the main factor influencing their relationship with other ethno‐
religious group in all parts of the country.
4.6 Muslims Reliance on Quran for Religious Guidance
Some of the statements of Muslim respondents indicate why they overwhelmingly rely on Quran for
religious figures. “Quran is the word of God; that is why we respect it”, “religious leaders also have to follow
Quran”, “Religious leaders cannot exist without Quran”, are some of the statements of Muslim respondents
interviewed in different localities.
High literacy rate even among Muslim women in Sri Lanka and the availability of Quran in local languages
have enabled the Muslim population in many parts of the country to rely on Quran for religious guidance.
Besides, Muslim religious leaders do not emerge as very prominent figures at a community level though
they play an important role in Muslim communities.
4.7 Practice of Islam: Why it is different in Sri Lanka?
Respondents recognize that the socio‐cultural environment in Sri Lanka has influenced the practice of Islam
in the county but they still feel that they have not deviated from the basic teachings of Islam. Muslims
continue to practice Islam according to the basic teachings of Islam guided by Quran and religious leaders.
On the other hand, Sri Lanka is not a Muslim country and its legal system is different compared to Islamic
countries.
33
Though certain Islamic traditions are dominant in the country, some Muslims have come under the
influence of minor local traditions specific to certain localities.
Conclusion
The qualitative study presented in the report was conducted in seven selected locations in the North and
East as well as in three sites in the rest of the country. These locations were selected on the assumption
that the presence of diverse religious communities there would affect inter‐community relations.
It is important to emphasize here that religious divisions in the country by and large follow ethnic divisions,
though there are exceptions in some areas. So, religious identity is often confused with ethnic identity. The
exceptions occur when the Sinhalese are divided into Buddhist and Christian faiths or when the Tamils are
split between Hindunism and Christianity. On the other hand, ethnic differences are often displayed by
respective ethnic groups in terms of religious practices, in particular, the public display of religiosity. This is
particularly so between the Sinhalese and the Muslims.
In most of the localities where the fieldwork was done, ethno‐religious communities tend to remain
physically segregated. For instance, many towns and villages are almost exclusively inhabited by one or
another community. This is particularly the case in Ampara, Baticaloa, Akurana, Kaluthara and Beruwala.
The members of ethno‐religious communities under such conditions tend to develop mutual suspicion.
Prevailing inter community tension at times leads to violent conflicts. Different religious groups belonging to the same ethnic group are not any better in overcoming mutual suspicion and inter community rivalries.
In Mannar, Tamils are divided between Hindus and Catholics. Their common ethnicity and the collective
experience of the adverse consequences of the war have not narrowed the gap between them.
Following the end of the conflict in the North and the East, people belonging to all communities are busy
settling down in their respective habitats. Those who left their villages to escape from violence are also
returning to their former habitats. This process is not smooth and often opens up all wounds when disputes
arise with regard to land rights and state patronage. Many people feel that deprivation and discrimination
follow religious and ethnic lines. This was particularly evident in Mannar and Batticaloa. Political leaders
who have come forward to facilitate the resettlement process often represent the interests of the members
of their respective communities. This has further reinforced ethno‐religious divisions.
In localities where ethno‐religious communities are not living in spatially segregated settlements, people
have tended to transcend narrow ethno‐religious divisions. Such harmony and co‐existence have been
facilitated by diverse circumstances such as the presence of far sighted community leaders, mutual
dependence between communities, sharing of common amenities etc. Yet, these are exceptions and in
most localities surveyed, competition, tension and mutual suspicion characterise the relations between
ethno‐religious communities. In spite of significant differences between ethno‐religious communities,
religious leaders have played a significant part to mediate between communities. Many people feel that it is
political leaders who keep the communities apart and contribute to disharmony.
It is in this context that people have greater faith in religious leaders, particularly at a community level.
Today it is common place for religious leaders representing different faiths to take part in community
events and this usually has a moderating influence on deep ethno‐religious divisions that characterize many
of the localities surveyed.
34
II. Methodological Report
1.1 Survey Period
The survey questions and structure were developed by The Asia Foundation and administered by Nielsen
Lanka Company (Pvt) Ltd between August 2 and September 23, 2011.
1.2 Sampling plan
The sample is comprised of the four major religious groups in Sri Lanka: Muslim, Buddhist, Hindu and
Roman Catholic2. The sampling frame is designed primarily to provide an understanding of each of these
religious communities at the national level, but also in selected provinces (section 1.5 below).
It is estimated that there are around 19 million people living in the nine provinces and 25 districts. Based on
the census data from 1981 and 2001, it is estimated that there are 12.4 million people aged 18 and older.
No accurate statistics are available for the Northern Province after the 1981 census. Estimations were used
for the purpose of this survey, however those figures are at the overall population level, and the gender,
ethnic, religious, and age breakdown were not adjusted. Additionally, demographic population data for the
Eastern Province is available to only a limited extent. Therefore, data on population percentage of religious
groups in the Northern and Eastern Province are based on census figures from 1981.
Population of religious groups by province (18 years and above, rounded)
Buddhist Hindu Islam
Roman
Catholic Sri Lankans
Western 2,603,000 183,000 282,000 416,000 3,483,000
Central 1,028,000 356,000 155,000 42,000 1,579,000
Southern 1,424,000 23,000 48,000 NA 1,494,000
North Western 1,054,000 30,000 155,000 179,000 1,415,000
North Central 656,000 8,000 59,000 NA 722,000
Uva 614,000 114,000 34,000 NA 760,000
Sabaragamuva 1,011,000 101,000 50,000 NA 1,161,000
Northern Province 20,000 639,000 44,000 116,000 818,000
Eastern 186,000 305,000 362,000 113,000 965,000
All Island 8,592,000 1,755,000 1,185,000 866,000 12,394,000
2 Roman Catholics were selected to represent the Christian community as whole because they constitute approximately
90% of the Christian population.
35
1.3 Target Population
Any Sri Lankan citizen living in households (own/rent, etc) in all the provinces during the survey time, aged
between 18‐70, and a follower of the four major religions—Buddhism, Hinduism, Islam or Roman
Catholicism—were defined as the target population for the survey. All nine provinces were included in this
survey, however in the Northern Province, parts of Kilinochchi and Mullaitivu Districts were not included
because of accessibility related to security issues. Individual persons were considered as the sampling unit.
1.4 Sample Size and Margins of Error
The total sample size of 5553 includes 2,070 respondents from the Muslim community, 1344 Buddhist
respondents, 1356 Hindu respondents, and 783 Roman Catholic respondents. The sampling is designed for
analysis at the national level for each of the four religious groups as well as in selected provinces.
As detailed in Table 2, the overall margin of error for the national sample is +/‐ 1.3% at 95% confidence
level. Margins of error for the four religious groups at the national level are: Muslims +/‐ 2.2%, Buddhists
and Hindus +/‐ 2.7%, and Catholics +/‐ 3.5%.
Table 2: Sample size and Margin of error by religious group
Population above
18 years Percentage Sample Size
Margin of
error %
Buddhist 8,592,000 69% 1344 2.7
Hindu 1,755,000 14% 1356 2.7
Muslim 1,185,000 10% 2070 2.2
Roman Catholic 866,000 7% 783 3.5
Total 4 groups 12,430,000 100% 5553 1.3
1.5 Sample Composition by Province
The sampling plan provides a wide spread across the country covering all 9 provinces for Muslims,
Buddhists, and Hindus. For Catholics, who constitute a relatively low percentage of the Sri Lankan
population, a group of provinces were targeted that contain over 80 percent of the total Catholic
population in the country. This was primarily done to avoid sensitivity associated with locating Catholic
respondents in provinces in which they are very small in numbers.
36
Table 3: Sample composition by religion across provinces
Province sample Buddhist Hindus Muslims Roman
Catholic
Total
Western 285 94 446 333 1158
Central 109 191 247 53 600
Southern 151 37 83 0 271
North Western 113 21 268 221 623
North Central 73 31 93 0 197
Uva 68 62 57 0 187
Sabaragamuva 113 62 68 0 243
Northern Province 31 429 403 130 993
Eastern 401 429 405 46 1281
All Island 1344 1356 2070 783 5553
Additionally, to allow for analysis of selected religious groups in certain provinces, oversamples were done
as follows: Muslims, Hindus, and Buddhists in the Eastern Province; Muslims and Hindus in the Northern
Province (Table 4). The survey sought to explore the Northern and Eastern Provinces in‐depth as these were
the provinces that were most severely affected by the war, and as shown in other nationwide surveys
harbor views and concerns unlike the rest of the country.
Table 4: Margins of error for oversampled religious groups in Northern and Eastern Province
Northern Province Eastern Province
Buddhists n/a 4.9%
Hindus 4.7% 4.7%
Muslims 4.9% 4.9%
1.6 Process of selecting sampling points and details of sampling point replacement
A stratified random sampling technique was used to select sampling points. The 5553 sample for each
religion was split by province and the province sample was further distributed across all districts. The
sample in each district again was split in order to cover the various Divisional Secretariats (DS). In each
DS division, Grama Niladhari (GN) divisions were selected as clusters using SPSS software random
number generation process.
37
In each GN Division, 10 interviews were targeted for the survey. GN Divisions were categorized based on
each religious population as follows:
1. 100‐81% population in one religious group
2. 80‐61% population in one religious group
3. 60‐41% population in one religious group
4. 40‐21% population in one religious group
5. less than 20% population in one religious group
The Department of Census and Statistics provides data on religious community populations by DS divisions.
This data was used in each province to select the DS divisions where the interviews were to be conducted.
The Grama Niladhari (GN) divisions within these DS divisions were then selected based on local knowledge
of settlements collected from key actors in the community. For each of the minority religious groups
(Muslims, Hindus and Catholics), areas with higher concentrations of their followers were sampled
wherever possible. However, if a small number of citizens were randomly selected in a GN in which their
religious group was low in population, interviews were still executed. For example, if in a Buddhist majority
sample point (GN) a few Muslim households were randomly selected, those interviews were carried out.
For 5553 interviews the total number of sampling points was designated at 556. However, because of
challenges that occurred during the fieldwork, problematic GN’s were replaced rather than risking further
problems and delays. Therefore, in some sampling points less than 10 interviews were conducted, resulting
in a necessary increase of sample points. The final total of sampling points was 563.
Sampling point replacement was necessary in a limited number of cases for two main reasons:
Due to the non‐availability/insufficiency of population figures for some areas, the target religious
population in a selected GN turns out to be limited.
Security issues and public disorder associated with the “greased devil”3 situation that took place
during the fieldwork period.
Additionally, a small number sampling point replacements were required in Jaffna (Northern Province) due
to it being a high security zone.
Table 5: Sample point distribution by province
Province Total
Central Province 64
Eastern Province 130
North Central Province 22
North Western Province 61
3 The “greased devil” situation refers to a series of mysterious sexual assaults and robberies of women that took place in several villages throughout Sri Lanka in August 2011. The acts were allegedly carried out by masked strangers covered in black paint, hence “greased devils”, and were believed to be coordinated. The bizarre incidents sparked widespread fear and a breakdown of law and order in several communities, marked by violence and clashes between citizens and police over response /lack of to the problem.
38
Northern Province 86
Sabaragamuwa Province 30
Southern Province 29
Uva Province 21
Western Province 120
Grand Total 563
1.7 Details of starting point selection
In each selected GN, 10 interviews were targeted for the survey and broken into two sets by starting the
interviews from two separate starting points to enhance the representativeness of the sample. Five
interviews were conducted from each starting point in every GN until 10 interviews were completed. In
each GN, easily identifiable landmarks were used as starting points, such as junctions, religious places,
schools, etc. Due to security threats associated with the Grease Devil situation, if problems arose in
sampling areas the interview was stopped and restarted from another starting point in the same GN or
revisited another day.
1.8 Household selection
The following procedure was adopted to select the households in each selected sampling point:
Household selection for interviews was done using the right hand walk method from a pre‐
determined starting point. All the households were approached until a successful interview was
completed. The religious quota for each GN was pre‐defined.
To ensure a spread in the area where the interviews were conducted, in urban areas every two
households were skipped and in a rural areas one household was skipped.
This process continued until five interviews were completed (10 interviews total from two starting
points)
In a few GNs, depending on the total sample size in the DS Division, less than 10 or more than 10
interviews were completed. For example, in some sample points a small number of additional
interviews were conducted as a result of the concentration of minority religious communities.
1.9 Respondent Selection Process
The survey was conducted in households with respondents above the age of 18 years from the specified
religious communities. The process of selecting a respondent from a household covered the following
aspects:
39
Introduction
Description of similar studies conducted and purpose of this survey
Identifying suitable members availability in the household
Selecting one respondent from the household randomly
The Kish grid was used for randomizing the target respondents within households. The interview continued
only if the respondent was willing to answer, otherwise the interview was terminated and the next
household was approached. Call‐back was allowed if the respondent was available to be interviewed later
on the same day in the same location.
2.0 GPS readings
GPS handsets were used by field staff to mark the coordinates of the 563 sampling points across the
country where the clusters of 10 interviews were conducted. GPS coordinates were not taken for each
household due to the likelihood of leading to refusal to participate in the interview. The GPS coordinates of
the sampling points serve to provide a reliable reading of the location and an indication of whether certain
viewpoints are geographically related.
2.1 Call back or substitution and percentage of refusals and non‐contacts
The survey was conducted during weekends, holidays and on days where most household members were
available at home. Call back and substitution were not allowed in the case of refusals and non‐contacts.
Table 6: Refusals and Non‐Contacts
Household refused to answer 11%
Individual selected person refused/ couldn’t answer 2%
Language issues 2%
Other reasons 2%
2.2 Quality Control
Overall 22 percent of the interviews were subject to quality control checks and 10 percent were
accompanied by supervisors from the field partner organization (Table 7). Interviews were back‐checked,
via the supervisor or other staff visiting the household after the interview was completed (typically two or
three days after) to check if the interviews had been conducted properly. The quality control checks verified
whether the random walk method was conducted properly, respondent selection using the Kish grid
was done correctly, and that no respondent substitution was made.
40
Table 7: Accompanied and Back Check
District
Total
Sample AC % BC %
Gampaha 393 8% 30%
Colombo 557 6% 19%
Ampara 304 20% 37%
Badulla/Monaragala 177 5% 29%
Anuradhapura/Vauniya 152 11% 34%
Kalutara 182 13% 24%
Kurunegala/Puttlam 401 5% 16%
Kegalle 100 25% 39%
Rathnapura 117 4% 26%
Polonnaruwa 119 23% 11%
Matara/Galle 204 13% 44%
Hambantota 49 35% 35%
Matale 85 45% 12%
Kandy 311 15% 19%
Neliya 182 14% 11%
Jaffna 298 7% 8%
Vavuniya/Puttalam/Kilinochi 583 4% 14%
Trinco 285 6% 29%
Batticaloa 371 15% 23%
Ampara 290 9% 24%
Colombo 100 0% 17%
5260 10% 21%
41
2.3 Weighting scheme
Two weighting schemes were used for the data:
Weighting scheme 1 was calculated using religion as the focus. The weighting factor was calculated by
working out the proportion of the number of people from a certain religious community from a particular
province as part of the total population of that same community nationally. For example, there are
2,603,000 Buddhists in the Western province. This forms 30% of the nation Buddhist population which is
8,592,000. The Buddhists from the sample in the Western province are given an appropriate weight to
make it of a similar proportion as the population. The data by religious communities, provinces, gender and
other segments are weighted by this particular weight scheme.
Weighting scheme 2 was calculated focusing on the population of the country as a whole. All sample
numbers achieved for a particular religious community in a province were brought at par with the actual
proportion of this religious group from a particular province in the national population. For example,
Buddhists in the Western province form about 21% of the national population (18 years and above). The
sample of Buddhists from the survey is given an appropriate weight to make it of a similar proportion as the
actual population. The national level numbers (‘All Island’4) have been weighted using weighting scheme 2.
4 It should be noted that “all island” includes the 4 major religious groups but not small segments such as other sects of Christianity
and minority religious groups that constitute a very low percentage of the population.
