Story by: Oxford Digital Gender Gaps team
On International Women’s Day, the Digital Gender Gaps team is excited to launch an updated version of our interactive dashboard. Our dashboard allows users to track internet use and mobile ownership by gender and gender gaps. We’re introducing some changes to our indicators in V2 of our dashboard.
New subnational indicators: we now feature a first-ever database of subnational indicators of internet use and mobile ownership by gender, and gender gaps, across 117 low- and middle-income countries (LMICs). The subnational data are administrative-level 1 geographies, and available from 2015 onward annually. From 2024, the estimates are at monthly resolution. These estimates are obtained from applying machine learning algorithms to social media, survey and geospatial data. We train models to predict digital access in contexts where we have data on key predictors of digital access – for example, measures such as Facebook penetration by gender, education, and income levels in a region – but don’t have direct measures from a survey on digital access. We calibrate and validate the performance of these models against data from the Demographic and Health surveys, for countries where these survey data are available. Through this approach, we can improve the geographical and temporal resolution of estimates of digital access by gender, as our method covers a broader set of countries and more recent time periods than available survey data.
Updated national indicators: Our national estimates continue to be derived with the same underlying methods as before, in which we compare and test against different predictive models relying on different variables or features to predict digital access measures. We continue to find that the best performing predictions come from a combined (online-offline) model that integrates social media indicators with population and development indicators (e.g. Human Development Index). In V2, however, we have switched to training the model largely on Demographic and Health Surveys and Multiple Indicator Cluster Survey estimates on digital adoption, with some additional estimates from the International Telecommunication (ITU) for high adoption only settings. We have also updated our predictive features. We have updated the national indicators to feature internet use and mobile ownership levels by gender, in addition to gender gaps. We also include trends over a longer period, with annual estimates now available from 2015 onward.
Tracking digital gender inequalities in a comparable way is a critical first step to ensuring that we can close the digital gender gap and ensure that women are equal participants in a digital society. Our work is one step towards that important effort, and we look forward to engaging with partners to use these data collectively.