For information about PAPI methodology and its development over time, see PAPI Methodology
How were the 30 provinces selected in PAPI 2010?: Covering roughly half the province’s number of the country, the 2010 study offered a rare opportunity to observe the effect of transparency brought by the index and its efficiency as a tool for policy making, by comparing the results of the participating provinces (the treatment group) to the provinces which don’t (the control group).
The method Propensity Score Matching was deployed to establish these two groups, identifying for each province in the treatment group a twin province in the control group, which is similar to it in factors with potential influence on governance and public administration, such as wealth, population size, urbanization, structure of the economy, etc. For more detailed information see PAPI Methodology