Construction of a New Index of Financial Inclusion

While we recognise that there exists no perfect method to construct indices in general, the problems with the widely used methods of constructing indices for financial inclusion (FI) stand out due to the peculiar nature of the variables of concern. First, FI is not unidimensional. It includes aspects of access, knowledge and usage of financial services at the least. Second, these different dimensions may not be perfect substitutes. Increasing financial literacy, for example, does not necessarily compensate for a lack of access due to the unavailability of bank branches nearby. Any index of FI thus must make sure that exceedingly good performance in one dimension should not be able to fully compensate for poor performance in another dimension in the index values. A discussion of how various researchers have tried to solve these problems in the past can be found in Sarma (2015) . We begin with a discussion of the Index of Financial Inclusion methodology proposed by Sarma (2015) and provide a new interpretation of what such indices may measure, based on a consideration of the distribution of the performances of various countries and the resulting entropy. Highlighting the unrealistic assumption of this methodology, we propose a new method to create sub-indices and the New Index of Financial Inclusion. JEL: C18, C40, G50, G59, O10