Forecasting socioeconomic trends with cell phone records

Abstract

National Statistical Institutes typically hire large numbers of enumerators to carry out periodic surveys regarding the socioeconomic status of a society. Such approach suffers from two drawbacks:(i) the survey process is expensive, especially for emerging countries that struggle with their budgets and (ii) the socioeconomic indicators are computed ex-post i.e., after socioeconomic changes have already happened. We propose the use of human behavioral patterns computed from calling records to predict future values of socioeconomic indicators. Our objective is to help institutions be able to forecast socioeconomic changes before they happen while reducing the number of surveys they need to compute. For that purpose, we explore a battery of different predictive approaches for time series and show that multivariate time-series models yield R-square values of up to 0.65 for certain socioeconomic indicators.

A survey of results on mobile phone datasets analysis

“…The authors propose several measures to quantify the mobility of users, and show that socio-economic levels present a linear correspondence with three indicators of mobility, namely the number of different antennas used, the radius of gyration and diameter of the area of often visited locations, indicating that the more mobile people are, the less poor the area in which they live seems to be. In a further study by the same research group, Frias-Martinez et al go one step further, and propose a method not only to estimate, but also to forecast future socio-economic levels, based on time series of different variables gathered from mobile phone data [63]. They show preliminary evidence that the socio-economic levels could follow a pattern, allowing for prediction with mobile phone data.…”

Relationships between mobile phone usage and activity-travel behavior: A review of the literature and an example

“…In addition to using socio-economic strata to explain the differences in mobile phone usage, researchers have done the other way around: using mobile phone usage to predict socio-economic information (Frias-Martinez et al, 2013;Soto et al, 2011). This kind of research is necessary as well because it is too expensive to carry out traditional census surveys nationwide, and at the same time, survey results are very likely to be soon outdated (Calabrese et al, 2013).…”

Computational Socioeconomics

“…Further, they proposed a model that can estimate the tower-level socioeconomic status with an adjusted R 2 = 0.72. Later, Frias-Martinez et al [491] analyzed a large-scale dataset of CDRs and suggested to predict future values of socioeconomic indicators based on human behavioral patterns. Using the multivariate regression analysis, they found that mobility variables perform better than consumption variables on predicting future socioeconomic indicators.…”