Incident Reporting: Experimental Data Collection Methods and Migration Governance
African migrations are transforming political power and authority on the continent. The growing presence of large populations of undocumented and/or disenfranchised people, particularly in urban centres, exacerbates and complicates the already tenuous relationship between ‘states’ and ‘citizens’ in this region. When holders of formal political authority have limited obligations to those that move through and reside in their jurisdiction, they are more inclined to reinvent their governance mandates and act outside the bounds of the law. Migrants also possess strong incentives to disengage from formal governance structures and in some cases, to deliberately subvert state agents, particularly those involved in immigration enforcement. Improved data on the informal relationships between state actors and migrants will help us to understand the evolving character of state sovereignty and territoriality in sub-Saharan Africa. Since informalisation influences the character and quality of official procedures for collecting migration data, this knowledge may also impact upon our capacity to develop reliable portraits of migration trends across the continent. Unfortunately, our ability to speak confidently about informality in migration governance is limited by a paucity of reliable and comparable data. The clandestine nature of many of the relevant activities, and the characteristic unreliability of individual testimonies compel us to conceptualise new approaches. While ethnographic studies offer potential ways around these problems, the acknowledged presence of an observer and the highly individualistic and idiosyncratic nature of this approach are constraining factors. This paper reviews the techniques and procedures that constitute, and the empirical and ethical strengths and limitations of, an experimental data collection method employed to correct this research gap in a study of street-level immigration policing in Johannesburg, South Africa. This approach, which we call ‘incident reporting’, combines a systematic procedure for sampling observed instances of immigration enforcement with a benchmarked process of categorising and coding these observations. The approach goes beyond conventional ethnography by a) decreasing, through the removal of threat of personal or institutional sanction and/or repercussion, incentives for subjects to adjust their behaviour or censor their language; b) increasing our capacity, through the utilisation of GIS mapping, to gauge the level of ‘disconnect’ between ordinary policing tactics and station-level plans; c) increasing our ability to generalise about a small set of observable facets of informal policing practices; and d) increasing the potential to compare informal practices across space and time. The paper discusses issues that require attention in order to transplant this approach to other research sites in Africa: a) the ethics of conducting clandestine forms of research; and b) the need to combine this approach with more conventional ethnographic study and key informant interviews.