services). As a result, reported crime rates often mask much higher instances of crime that goes unreported. As Ashwani Saith notes:
While the conceptual exercise travels … from values to indicators, the operational one then has to follow in the reverse direction: starting with numerically specified indicators that are successively combined and aggregated to eventually emerge with an overall quantitative reading on the extent to which the desired values were realized. At the end, it is the practicality and availability of data that sets the real binding constraint on which indicators get used and which not, and often conceptually stronger specifications have to yield to inferior ones for lack of reliable data (2006: 1183-1184).
Gilgen and others note these challenges, writing in support of the armed violence proposal, recognizing the lack of credible data in many African countries, as well as the problems of low reporting or non-use of services (2010: 16). Yet they point to improvements in data collection to date, and the importance of redoubling these efforts to ensure that the best indicators can be used, rather than relying on less useful indicators simply due to issues of data availability (Gilgen et al 2010: 22).
It is also important not to overstate the challenges of data collection. While it is a critical factor limiting the effectiveness of measurements, particularly in fragile states, the MDG process was a good illustration of how data collection problems can be overcome when these is sufficient political will. If countries decide that a particular issue is sufficiently important to development prospects, new data can be collected. While this will not be without a multitude of methodological challenges, it nonetheless highlights that the data collection problems are not necessarily insurmountable