something as straightforward as homicide, do you measure by rates of reporting to police, prosecutions or convictions? Each of these depends upon a host of factors that have the potential to influence the data – such as the likelihood or culture of reporting crimes, the institutional capacity of the police and the courts, etc. As Vandemoortele points out, ‘all indicators are imperfect but some are more imperfect than others’ (2008: 222).
Challenges of measurement are particularly striking in fragile environments, where data collection can be especially difficult. First of all, record keeping within the police, courts and prisons (as well as other services) is likely to be poor, making it difficult to use national data for measurement purposes. Furthermore, institutional capacity constraints can make even data that is available of little use. For example, if there is no functioning court system to prosecute individuals or a weak police service to arrest and detain suspected criminals, crime rates are likely to appear artificially low. And finally, there is very often not a culture of reporting crimes in fragile states. This is due in part to cultural taboos, as well as a lack of trust in state institutions (particularly the security