Perhaps the most important barrier to doing survey work in Karachi is also the one that is noticed the least. Our survey like most other surveys in Pakistan took gender segregation in to account and ensured that women enumerators would interact with female respondents. It is good practice everywhere for women and girls to be interviewed by female surveyors, particularly when the topics discussed might include sensitive issues relating to reproductive health or abuse. The segregation that exists in rural communities is also strongly present in Karachi. Survey design needs to assume that it will be impossible for our male team members to speak with women and older girls in the community. Not surprisingly, this was more or less borne out.
Other issues in access, however, did surprise me. Although ethnic identities in Karachi are very pronounced, issues in data collection were the same across various groups. Irrespective of ethnic identity people of similar socio-economic backgrounds are confronted with more or less the same sorts of problems in their daily lives. Hence, while mass migration to the city in the last 20 years has divided people into different groups and communities, similar problems have brought them together and made them react at times in the same way. In a few areas ethnic differences have resulted in conflict, but this is mainly due to affiliations with rival political parties. In these areas it was necessary to ensure that fieldworkers spoke the ethnic language of the respondents to avoid creating animosity within the community.
Gaining access to localities was one of the most challenging problems faced during the exercise. Communities restricted access to outsiders due to a number of reasons. Some of these can be attributed to the deteriorating law and order condition, strong political rivalries and the presence of armed groups. It became imperative for the survey teams to have prior knowledge of these issues, making it important to gather a profile of each community, highlighting sensitive localities. Working with resource people from these areas to help facilitate the research proved to be a useful strategy. In fact there were some areas where the exercise was entirely dependent on the support of local resource persons.
Some localities in District West and District South were more difficult to access due to the targeted operation going on there against religious extremists (supporters of Tehrik-i-Taliban Pakistan) and other armed groups. The strong influence of religious organization in some localities of District West also aggravated the situation. In these communities there was a sense that any association with international donor agencies was somehow un-Islamic. Even in these places, however, local resource persons who understood the reasons for our study proved to be useful bridges to the community.
Hesitation on the part of some respondents to participate in the survey was telling of their perceptions of the development sector and in general reflected a cautious attitude towards outsiders. One reason why people were initially unwilling to take part in the exercise is that the prevalence of crime has resulted in a trust deficit. This situation was more evident in middle income groups in District Central where respondents raised questions about the organization’s affiliation and needed to be convinced about the purpose of the survey. People were also suspicious of NGOs and related the Collective’s research to NGO work and hence were reluctant to participate.
Data collection also revealed the extensive reach of the national cash transfer programme, the Benazir Income Support Programme (BISP). Many people had participated in the BISP’s Poverty Score Card survey to determine their eligibility, and whether they were enrolled into the programme or not often had a bearing on their attitude towards participating in the Collective’s survey. The surveyed households who had gone through the lengthy process of determining their eligibility for the BISP and had not become beneficiaries of the programme showed a lack of interest during the survey and complained of the futility of these exercises.
In the end, we were able to navigate the obstacles that come in the way of doing survey work in a complex city like Karachi. Issues such as ethnic divisions, the presence of armed groups, religious and political mistrust, fear of crime, disappointment with past surveys, and the ongoing security operation are all important factors in hindering access to communities, families and individuals. While these barriers seem insurmountable when seen from a distance, we found that common sense measures such as working close to the ground, using local resource persons, remaining sensitive to local conditions and sensitivities, and being flexible in implementing our work schedule, open up all types of localities. Paradoxically, we had the greatest difficulty in gaining access to individuals in relatively better off and educated neighbourhoods where communities were extremely cautious with respect to security and individuals were most concerned about their privacy.
Despite its many challenges, much of Karachi remains open to survey most of the time.