Learning Center
Plans & pricing Sign in
Sign Out

GBV Pakistan Survey Results FINAL


									Gender-Based Violence among Afghan Refugees
Summary of Post-intervention Survey Findings in Three Camps in Northwest
Frontier Province, Pakistan

International Medical Corps Pakistan – January 2010

INTRODUCTION                                                   as 79%,1 suggesting that underreporting could have
Violence against women and girls is a major public             affected the survey results.
health and human rights issue worldwide, and is
often exacerbated by the stress and disruption of              Even after International Medical Corps’ GBV
people’s lives in conflict-affected settings, such as          program, most women reported that they did not
Pakistan and Afghanistan. Data regarding the current           tell anyone about their abuse mostly because they
nature and extent of gender-based violence (GBV)               believed the violence to be normal and speaking out
among Afghan refugees living in Pakistan are very              would not change their circumstances. While this
limited. In September 2008, International Medical              suggests that there is little empowerment among the
Corps initiated a comprehensive, community-based               women interviewed, the post-intervention survey
GBV prevention and response initiative in four                 found an 11-point increase in female respondents’
Afghan refugee camps (Koga, Barakai-I, Barakai-II,             knowledge of their rights under Islam and a 32-point
and Mera Kachawri). The objective of the one-year              increase in their understanding that they can refuse
program, funded by the U.S. State Department                   sex with their husband. There was also a significant
Bureau of Population, Refugees, and Migration and in           decrease in the number of women and men who
close coordination with the UN High Commissioner               view wife beating as an acceptable way for husbands
for Refugees (UNHCR), was to raise awareness                   to discipline their wives. The increase in knowledge
about GBV, reduce the incidence of GBV, and                    and understanding of women’s rights, as well as the
improve the health of GBV survivors. As part of this           change in perceptions against GBV, suggests that the
GBV initiative, International Medical Corps carried            communities are becoming increasingly exposed to
out a baseline survey in November 2008 to assess               anti-GBV messages. However, much remains to be
male and female community members’ knowledge                   done to decrease the occurrence of GBV within
and attitudes regarding GBV. A follow-up survey was            Afghan refugee camps, as more than half of the
then conducted in August 2009 to evaluate the                  population reports some form of violence within
impact of the GBV program on the targeted                      their households.
                                                               SURVEY METHODS
SUMMARY OF FINDINGS                                            A total representative sample of 260 households was
The post-intervention survey found GBV to be highly            randomly selected from updated household lists of
prevalent among those interviewed, with more than              each of the four camps. Two camps (Barakai I and II;
50% of women reporting to have experienced                     65 per camp) were randomly assigned for interviews
physical or emotional violence. Sixty-percent of men           with men only and the other two camps (Mera
reported that they have abused their wives in some             Kachawri and Koga) for interviews with women
form, with 30% admitting to more severe forms of               only. To select participants at the household level,
abuse, such as punching with their fist and other              International Medical Corps-trained community
actions that could cause serious harm. Emotional               health workers (CHWs) visited the selected
abuse was reported among 32% of women                          households and completed a grid to randomly select
interviewed, mostly in the form of withholding                 one eligible, married male or female (depending on
money or humiliating them in front of others.                  the camp) of reproductive age (15-49 years).
However, while these prevalence rates are high,
similar studies have found abuse rates to be as high           1
                                                                 Ward, J. (2002). If not now, when? Addressing gender-based violence
                                                               in refugee, internally displaced, and post-conflict settings. T. R. H. f. R.
                                                               Consortium, The Reproductive Health for Refugees Consortium.

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 1 of 10
Interviews themselves were conducted at the Basic                      household was interviewed. This gave an initial total
Health Units (BHUs) or another secure location. To                     sample size of 260 households, with 130 male
ensure     privacy      and     protect     respondent                 respondents and 130 female respondents. However,
confidentiality, International Medical Corps GBV staff                 as mentioned above, insecurity reduced the female
conducted all interviews, with International Medical                   survey size by half, as a follow-up survey in the
Corps     male      counselors     interviewing   male                 fourth camp, Koga, was not possible. Out of the total
respondents and female counselors and trainers                         sample size, some people chose not to participate,
interviewing       female      respondents.     Survey                 leaving a survey sample size of 122 males and 62
questionnaires for men and women were developed                        females for the pre-intervention survey and 125
based on the Compendium of Monitoring and                              males and 60 females for the post-intervention
Evaluation Indicators for Violence against Women                       survey.
and Girls (VAW/G), the CDC Reproductive Health
in Crisis Toolkit, and the DHS Domestic Violence                       The respondents from the follow-up survey are not
Module.2                                                               the same individuals as those interviewed for the
                                                                       pre-intervention survey; however, they were
Following the GBV prevention activities implemented                    selected in a similar manner. Males and females were
by International Medical Corps, the CHWs returned                      sampled from different camps as recommended for
to Barakai 1, Barakai II, and Mera Kachawri and                        respondent protection purposes.
conducted a follow-up survey, which was carried out
in the same manner as the baseline pre-intervention                    DEMOGRAPHIC CHARACTERISTICS
survey. Unfortunately, due to military operations in                   The average age of male and female respondents in
Buner District during the summer, the CHWs were                        the pre-survey was 31 and 30.5 years old,
unable to return to Koga camp for the follow-up                        respectively. The average age in the post-survey was
survey. Therefore, the baseline survey results for                     29 years old for male and 31 years old for female
Koga camp were not included in this report.                            respondents. The majority of male respondents, pre-
                                                                       survey (89%) and post-survey (100%), were ethnically
SURVEY SITES                                                           Pashtun. However, only 44% of women who were
Barakai-II refugee camp is approximately 113 km                        interviewed for the pre-survey and 62% who were
from Peshawar in Swabi District. The total                             interviewed for the post-survey were Pashtun3. The
population of the camp is 24,086, with 5,298 women                     average age that the respondents were first married
of childbearing age. Most families originated from the                 was 20 for males and 17 for females.
eastern region of Afghanistan and are ethnically
Pashtun. Barakai–I refugee camp, also in Swabi
District, has a population of 25,155, with 5,534
women of childbearing age. The families are mainly
from eastern Afghanistan and are also ethnically
Pashtun. Mera Kachawri is located in Peshawar and
has a population of 25,105 individuals, with 5,523
women of child-bearing age. Like with the other
camps, residents of Mera Kachawri are also originally
from eastern Afghanistan.

Within each camp, 65 households were selected
through random sampling, and one person per

 Bloom, S. Violence against Women and Girls: A Compendium of 
Monitoring and Evaluation Indicators. Measure Evaluation. October 
2008. MS‐08‐30; Center for Disease Control. Reproductive Health in 
Crises: Reproductive Health Assessment (RHA) Toolkit for Conflict‐
Affected Women. January 2007; Measure Evaluation. Domestic Violence    3
                                                                        The options for Ethnicity were Pashtun, Tajik, or Other. Most who did
Module. September 2005.                                                not pick Pashtun chose Other.

    Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
    Page 2 of 10
Table 1. Educational Background of Respondents
Table 1. Educational Background of Respondents
                                                                         Male                     Female
                                                                         Pre         Post
                                                                                     Post         Pre          Post
Highest level of education completed
Highest level of education completed
  None                                                                   43%
                                                                         43%         26%
                                                                                     26%          76%
                                                                                                  76%          85%
  Primary                                                                25%
                                                                         25%         25%
                                                                                     25%          3%
                                                                                                  3%           0%
  Secondary                                                              10%
                                                                         10%         21%
                                                                                     21%          0%
                                                                                                  0%           0%
  College                                                                14%
                                                                         14%         21%
                                                                                     21%          0%
                                                                                                  0%           0%
  University                                                             1%
                                                                         1%          3%
                                                                                     3%           0%
                                                                                                  0%           2%
  Religious                                                              7%
                                                                         7%          4%
                                                                                     4%           18%
                                                                                                  18%          13%
Reading Ability
Reading Ability
  Can read easily
  Can read easily                                                        42%
                                                                         42%         61%
                                                                                     61%          2%
                                                                                                  2%           8%
  Can read with difficulty
  Can read with difficulty                                               12%
                                                                         12%         16%
                                                                                     16%          21%
                                                                                                  21%          5%
   Cannot read
   Cannot read                                                            46%
                                                                          46%          23%
                                                                                       23%            77%
                                                                                                      77%        87%
*The percentages respondents who chose not not to answer questions presented in this table are not included
*The percentages ofof respondents who chose to answer these these questions presented in this table are not included

SURVEY LIMITATIONS                                                     be noted, however, that while there are some
As with other pre- and post-intervention surveys,                      ethnic differences, all of the residents of the
this study has some limitations, which include:                        camps have been living within the camps for
• The duration of the campaign was too short to                        close to 25 years. Therefore, cultural beliefs and
     fully assess the level of uptake of GBV                           practices tend to be similar across families.
     prevention messaging in the community and see                 •   It is possible that underreporting of intimate
     the behavioral effects of the knowledge acquired.                 partner violence occurred due to social stigma
     As with any program that focuses on behavior                      and/or fear of reprisal. As common with other
     change, more than one year is required for the                    GBV programs, the taboo surrounding this topic
     community to truly internalize the information                    tends to inhibit respondents from answering
     presented and demonstrate true change.                            truthfully. Also, surveys in general must take into
• The small sample size might not be truly                             account social desirability bias, i.e., when the
     representative of the entire population of the                    interviewee provides answers based upon what
     four camps. The limited sample size was due to                    s/he thinks the interviewer wants to hear. As this
     several factors. The first was the difficulty in                  study relies solely on data collected from these
     recruiting enough people, particularly females,                   surveys, it is difficult to ascertain the true impact
     who were willing to be interviewed on such a                      on the incidence of GBV.
     sensitive subject. The second was the time that it
     took to perform the interviews. Also, the
     security situation prevented International                    INTERNATIONAL MEDICAL CORPS
     Medical Corps from performing, the follow-up                  ACTIVITIES
     survey in Koga camp, which reduced the number                 International Medical Corps worked closely with the
     of female respondents by one half.                            community and community leaders to develop lesson
• There were differences in the background and                     plans covering topics related to GBV and human
     demographic characteristics of respondents of                 rights. These lessons defined physical, psychological,
     the baseline versus the follow-up survey. While               and sexual violence and covered topics such as basic
     respondents were chosen in a similar manner                   human rights, child rights, women’s rights, and
     and from the same three camps, characteristics                refugee rights. Developed in close coordination with
     such as education level differed slightly. It should          the community’s religious leaders, most of the

 Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
 Page 3 of 10
messages were rooted in Islamic teachings,                     received GBV prevention education while they were
frequently referencing the Qu’ran. Lessons also                visiting a facilitiy. International Medical Corps-trained
emphasized the physical as well as the mental health           CHWs and TBAs conducted an average of 124
implications of GBV, not only for the victim, but also         awareness sessions per month in the target camps
for other household members who witness the                    for the communities. IEC material on GBV was
violence, such as children.                                    developed in collaboration with the community
                                                               representatives in the local language and 7,610
International Medical Corps counselors delivered               copies were distributed among the target population.
these lessons at monthly awareness sessions for each
camp’s male and female health committees. These                In addition to raising awareness, International
lessons were also delivered to patients and attendees          Medical Corps trained CHWs, TBAs, doctors, and
visiting the International Medical Corps health                other medical staff in the identification, counseling,
facilities on a weekly basis. Larger seminars targeting        and medical management of GBV cases. Before the
teachers, religious scholars, health committee                 project, no GBV survivors were reported in the
members, and community elders were also                        target camps. During the project implementation, 24
organized, where these topics were discussed at                GBV survivors were reported and were provided
length      and     information,     education,    and         medical and psychosocial support as required.
communcation (IEC) materials were presented and                International Medical Corps also translated the
approved       by    community     leaders.    Further,        UNHCR health care referral form into the local
International Medical Corps trained CHWs and                   language, Pushto, and devised a referral system in
traditional birth attendants (TBAs) to deliver these           collaboration with UNHCR for GBV survivors to
messages throughout the community.                             receive additional care.

Overall, a total of nine seminars for males and 11             RESULTS
seminars for females were held, in which 1,318                 Nearly half of the women interviewed reported
individuals participated. This is in addition to the 87%       having experienced physical or emotional violence
of health committee members who took part in the               from their husbands (Table 3). The most common
monthly awareness sessions and the 57% of patients             forms of violence reported by women were slapping
of International Medical Corps’ health facilities who          and having objects thrown at them (Figure 2). The

 Table 2. Experience with intimate partner violence reported by women and men
 Indicator                                                                              Pre Survey       Post Survey
 Reported by Women
 Proportion of married women aged 15-49 who ever experienced physical violence          49%              48%
 from their husband+

 Proportion of married women aged 15-49 who have ever experienced emotional             48%              47%
 violence from their husband+
 Proportion of married women aged 15-49 who ever experienced sexual violence            14%              10%
 from their husband+
 Reported by Men
 Proportion of married men aged 15-49 who reported ever being physically violent        69%              64%
 toward their wife
 Proportion of married men aged 15-49 who reported ever being sexually violent          6%               6%
 toward their wife
 +-Refers to indicator derived from the 2008 Compendium of M&E Indicators for VAW/G

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 4 of 10
most common forms of emotional violence reported               given that women within these communities have
were withholding of money or access to money                   been raised with the notion that they are not to
(32%), saying or doing something to humiliate them             discuss “household matters” with outsiders;
in front of others (28%), and being insulted (12%).            therefore, this deeply entrenched behavior will take
                                                               time to change. The difference between the number
More than 60% of men reported having physically                of men who admit to violence and the number of
abused their wives in some form, most frequently               women who reported ever having experienced
through pushing and slapping (Figure 3). About 30%             violence suggests that the taboos surrounding this
reported perpetrating more serious intimate partner            type of discussion remain stronger for women than
violence (IPV) against their wives, such as punching           for men.
with a fist or inflicting serious physical harm. Sexual
violence was the least common form of IPV reported             Approximately 12% of women reported that they
by both men and women, but it was also the form of             had experienced some form of IPV within the last 12
violence most likely to be underreported.                      months, compared with 40% in the baseline. These
                                                               women were asked about any injuries and coping
The fact that there was no change in the proportion            mechanisms. More than half of them reported
of women who responded that they had experienced               injuries such as cuts, deep wounds, eye injuries,
violence may indicate that women continue to feel              sprains, or dislocations. As with the baseline study,
uncomfortable discussing IPV. This is not surprising           most of the women who experienced violence did

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 5 of 10
Table 3. Respondent knowledge and attitudes toward women’s rights and GBV
Indicator                                                                    Male                  Female
                                                                             Pre        Post       Pre    Post
Proportion of people who know any of the rights of women according to
Islam                                                                        94%        94%        79%    90%
Proportion of people who agree that a husband is justified in beating his
                                                                             70%        52%        91%    50%
Proportion of people who would offer assistance to a woman who had
                                                                             48%        60%        84%    83%
been beaten by her husband
Proportion of people who say that men cannot be held responsible for
                                                                             66%        58%        41%    23%
controlling their sexual behavior+
Proportion of people who agree that a woman has a right to refuse sex
                                                                             95%        95%        65%    97%
with her husband+
Proportion of people who agree that rape can take place between a man
                                                                             62%        47%        62%    38%
and woman who are married+
+Refers to indicator derived from the 2009 Post Survey for VAW/G

The main reasons women gave for not talking to                  The study revealed that a majority of people, both
anyone about the violence they experienced were:                before and after program implementation, were
(1) belief that violence is normal and there is no              aware of the rights of women under Islam.
benefit to complaining, (2) fear it would bring a bad           Respondents were asked if they were aware of
name to the family, and (3) belief that it would cause          women’s right to work, marry whom they want,
more problems in the relationship. When asked                   divorce, own property, and vote. There was a
whether there were steps that might help them in                significant increase in the number of women who
coping with their experiences of violence, the most             responded that they knew women had the right to
common answer they gave was “talking with                       divorce (55% baseline to 75% follow-up) and to vote
family/friend.” The second most common answer                   (44% baseline to 73% follow-up) under Islam.
was “trying to forget about it.”

 Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
 Page 6 of 10
Table 4. Respondents’ attitudes toward gender-based violence
Indicator                                                    Male                               Female
                                                             Pre             Post               Pre               Post
                                                             Survey          Survey             Survey            Survey
Proportion of people who agree that women should have
the same rights as men                                       84%             78%                95%               88%

Proportion of people with gender-related norms that put
women and girls at risk for physical and sexual violence+    96%             72%                49%               31%

Proportion of people who believe that men can prevent
                                                               69%           53%                58%               58%
physical and sexual violence against women and girls
+Refers to indicator derived from the 2009 Post Survey for VAW/G

                                                                follow-up survey saw a dramatic increase in the
There was a significant drop in the proportion of               number of women who stated that women have the
both men and women who felt that wife beating is an             right to refuse sex with their husbands (Table 2).
acceptable way for husbands to discipline their wives.
As illustrated in Figure 4, the decrease was                    There was a decrease in the proportion of people
experienced across all five proposed justifications for         who agree that rape can occur between a man and a
both men and women; however the change was                      woman who are married (Table 2). This
more dramatic for women, who started off at higher              demonstrates that people within these communities
levels of acceptance of wife beating during the                 continue to associate rape as an act that happens
baseline study.                                                 outside of marriage, and that sex with a spouse,
                                                                whether forced or not, is legitimate.
A high proportion of people, mainly women,
responded that they would assist a woman being                  In terms of male attitudes toward GBV, the
beaten by her husband. While this may indicate a                proportion of men who agree that women should
lower acceptance level of intimate partner violence             have the same rights as men decreased in the follow-
in the community, it should also be noted that this             up survey (Table 3). There is also a decrease in the
question is prone to social desirability bias, in that          number of men who believe that men can prevent
respondents may be giving answers that they feel are            physical and sexual violence against women. While
acceptable. Of those respondents who said they                  these two findings may be due to the limitations of
would not assist, the primary reason given was that it          the survey as explained above, they should also be
was a private matter between a husband and his wife.            taken as an indication that more needs to be done in
                                                                terms of GBV prevention messaging and education.
There was a decrease in the proportion of people
who felt that men could not be held responsible for             In measuring the proportion of men and women who
what they do sexually (Table 2). While again this               hold attitudes about gender norms that reinforce
question may be prone to social desirability bias, it           violence against women and girls, statements were
may also indicate that there has been a positive                read and the respondent was asked if they agreed or
change in norms and attitudes at the community                  disagreed.4 The norm statement with which
                                                                  These statements were taken from the Gender-Equitable Men (GEM)
A high proportion of men responded in both surveys              scale that was developed by Pulerwitz and Barkerto to measure the
                                                                impact of an intervention on changing attitudes toward gender-related
that a woman has the right to refuse sex with her               norms. The Violence Against Women and Girls (VAW/G) Compendium of
husband if she does not desire it (74%), if she is              Monitoring and Evaluation Indicators has developed an indicator based on
feeling ill (92%), and/or she is pregnant (94%). The            GEM but focusing exclusively on VAW/G.

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 7 of 10
respondents agreed the most was “a woman should                to messages about violence against women is due to
tolerate violence to keep her family together.” There          the differences in exposure by camp, it is more likely
was a significant decrease in the number of men who            accounted for by the relative isolation of women,
agreed with the statement that it is a man’s right to          their limited mobility, and high levels of illiteracy
beat a woman, and a large drop in the number of                compared to men. Addressing this gender-based
women who responded affirmatively to the                       disparity in access to information will require
statement that there are times when a woman                    innovative, community-based approaches.
deserves to be beaten. While this could indicate a
shift in beliefs, it could also be explained by greater
exposure to the GBV prevention messaging.

About the same proportion of men and women                     CONCLUSIONS AND
reported that they had heard about issues related to           RECOMMENDATIONS
violence against women and girls through local print,          Gender-based violence continues to be a reality that
television, or radio media (70% of men compared                Afghan refugee women live with everyday. In
with 68% of women). A higher proportion of men                 addition to a prevalence of beliefs and attitudes that
than women reported having talked about GBV-                   put many Afghan women at risk of violence, studies
related issues with someone who came to their                  have shown that refugee settings can lead to an
house or their neighborhood in the past six months             increase in family tension due to poor living
(73% of men compared with 57% of women). The                   conditions, high rates of unemployment, and
dramatic increase in percentages for female                    overcrowding in camps.5 Over the past year,
respondents who had been exposed to prevention                 International Medical Corps’ GBV program has
messages may indicate that the program has been                carried out an awareness campaign aimed at
successful in reaching women in this target                    educating the community on their rights as well as
population.                                                    the negative effects of GBV. The purpose of this
                                                               study was to measure the effectiveness of this
Knowledge of available services among women had                campaign by comparing results from a pre-
also increased, with 35% of respondents naming                 intervention      and       a        post-intervention
without any help one or more providers delivering              knowledge/attitudes/practice survey.
services for those affected by violence, such as a
BHU, UNHCR, or the Commission for Afghan
Refugees. The most common provider cited was the               5
                                                                 Adnan A. Hyder, Zarin Noor, Emma Tsui (2007). Intimate partner 
BHU for both male and female respondents. While it             violence among Afghan women living in refugee camps in Pakistan.
is possible that the gender discrepancy in exposure            Social Science & Medicine 64 (2007) 1536–1547. 

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 8 of 10
 Table 5. Exposure to GBV prevention messages and knowledge of available services
 Indicator                                                                   Male                  Female
                                                                             Pre         Post      Pre      Post
 Proportion of people who have been exposed to GBV prevention                52%         70%       3%       68%
 messages in the last six months through media+
 Proportion of people who have been exposed to GBV prevention                50%         73%       11%      57%
 messages in the last six months through in-person communication+
 Proportion of people who demonstrate knowledge of available social          44%         66%       5%       35%
 welfare-based GBV services+
 +-Refers to indicator derived from the 2008 Compendium of M&E Indicators for VAW/G

                                                               in the camps. While awareness raising is a good first
The study had several limitations, including: the              step, the community still needs to internalize these
limited timeframe of both the study and the overall            messages in order to demonstrate real change.
                                                               There are also some messages that may not have
program, difficulties in establishing a statistically          been fully communicated, as evidenced by the high
significant sample size, potential differences in              proportion of respondents who believe that rape
demographics of pre- and post-survey respondents,              cannot occur between husband and wife and that
and the potential for underreporting due to the                men cannot prevent physical and sexual violence
sensitive nature of the survey topic. Nevertheless,            against women and girls. Further, a strong stigma still
despite its limitations, the study provides insight into       inhibits women from discussing the issue and from
some of the positive changes in knowledge and                  seeking help if it is required.
attitudes related to human rights and GBV that have
occurred over the past year, as well as areas where            By continuing to work with the community to deliver
changes have yet to occur.                                     GBV prevention messages in a culturally sensitive
                                                               manner, International Medical Corps can continue to
One of the most dramatic changes highlighted by the            support the protection of women and girls and
study is the increase in knowledge of women’s rights,          improve the health of the Afghan refugee
most significantly by women. Post-intervention                 community.
interviews show an 11-point increase in female
respondents’ knowledge of women’s rights under
Islam and a 32-point increase in their knowledge that
a woman has the right to refuse sex with her
husband. This increase is most likely tied to the
upsurge of all respondents who have been exposed
to GBV prevention messages over the past year
(18% increase for men; 65% increase for women
who have been exposed to GBV messages through
media). There was also a significant decrease in both
women and men who demonstrate gender-norm
related attitudes that reinforce violence and who                    International Medical Corps is a global,
responded that wife beating is an acceptable way for                 humanitarian, nonprofit organization dedicated
husbands to discipline their wives.                                  to saving lives and relieving suffering through
                                                                     health care training and relief and development
However, the study also showed that much more                        programs.
must be done to truly decrease the incidence of GBV

Knowledge, Attitudes and Exposure to GBV among Afghan Refugees in Pakistan, Final Report December 2009
Page 9 of 10

To top