Can Improvements to Mobile Internet Service Help Reduce Digital Inequality? An Empirical Analysis of Education and Overall Data Consumption
Management Science · 2025 · Karthik Babu Nattamai Kannan、Eric Overby、Sridhar Narasimhan
中文摘要
优质的互联网接入对于参与当代社会至关重要。遗憾的是，许多家庭——尤其是社会经济地位较低的家庭和/或农村地区家庭——无法获得优质的互联网接入。一些家庭完全没有接入，而另一些则依赖其移动数据套餐上网（即“智能手机依赖”）。这造成了互联网接入与使用方面的不平等。鉴于许多弱势家庭依赖智能手机，我们探讨移动互联网服务的改善能否有助于缩小数字不平等。我们聚焦于一项具体的改善：无限移动数据的获取。要使无限数据的获取有助于缩小数字不平等，它必须为弱势家庭带来比优势家庭更大的数据消费增长，包括对可能提升福利的数据（如在线教育内容）的消费增长。这一结果并非显而易见。据此，我们使用一家大型电信公司的详细订户层面数据，考察订户在改用无限移动数据套餐后对教育及其他内容的消费变化。我们发现，尽管所有订户的消费均有所增加，但弱势订户的增幅显著更大，无论是整体内容还是教育内容。鉴于识别出能为弱势家庭带来不成比例的数据消费增长（包括对教育及其他“提升型”内容的消费增长）的项目是缩小数字不平等的必要一步，这一发现具有重要意义。
Abstract
Quality internet access is critical to participating in contemporary society. Unfortunately, many households—particularly those of low socioeconomic status and/or those in rural areas—do not have quality internet access. Some have no access, whereas others are reliant on their mobile data plans for internet access (i.e., they are “smartphone dependent”). This generates inequality in internet access and use. Given the smartphone dependence of many disadvantaged households, we explore whether improvements to mobile internet service can help reduce digital inequality. We focus on a specific improvement: access to unlimited mobile data. For access to unlimited data to help reduce digital inequality, it must generate larger data consumption increases for disadvantaged households than for advantaged ones, including for data likely to enhance welfare, such as online education content. It is not obvious that this will be the case. Accordingly, we use detailed subscriber-level data from a major telecommunications company to examine changes in the consumption of education and other content after subscribers switch to unlimited mobile data plans. We find that although all subscribers increase their consumption, the increases are significantly larger for disadvantaged subscribers, both for overall content and for education content. This is an important finding given that identifying programs that generate disproportionate data consumption increases for disadvantaged households—including for education and other “enhancing” content—is a necessary step for reducing digital inequality.