Mobile phone price as a proxy for socio-economic indicators

Average mobile phone price of a region's residents has not been explored as a proxy variable for socio-economic indicators in development literature. Using handset models from mobile data logs of a telecom operator, we calculate average phone prices of subscribers at each cell tower, and then find their relationship with indicators from preexisting census data. Our results show that average phone price correlates well (absolute Pearson's correlation coefficient > 0.35 and p-value < 0.01) with indicators of accessibility to services, infrastructure, hygiene and communication, but does not correlate well with literacy variables.