https://pulse.internetsociety.org/blog/the-new-digital-divide The New Digital Divide [image: Picture of Shane Greenstein] Shane Greenstein <https://pulse.internetsociety.org/blog?author=127> Guest Author | Harvard Business School <https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fpulse.internetsociety.org%2Fblog%2Fthe-new-digital-divide> July 15, 2025 In short - A new study offers an unprecedentedly granular look at digital usage in the United States, including how people browse the Internet, use social media, create documents, and code software. - Areas with higher household incomes and greater educational attainment consistently showed higher digital engagement. - These findings challenge the assumption that digital divides disappear once basic broadband infrastructure is widely available. The digital divide is typically understood as the gap between those with access to Internet infrastructure and those without. My colleague’s recent research, though, reveals that the real story goes deeper, highlighting stark differences in how digital technologies are used. This ‘new digital divide’ has significant consequences for social equality, as it exacerbates existing disparities and creates new ones. It also alters economic opportunities, as those with limited digital skills may be left behind in the job market. Furthermore, it affects civic participation, as those who are less digitally literate may struggle to engage in online civic activities. A Substantial Divide Exists Between Urban and Rural Communities Our research (see paper <https://ssrn.com/abstract=4957386>), conducted in partnership between the Microsoft AI for Good Lab, Harvard Business School, and the Wharton School, offers an unprecedentedly granular look at digital usage in the United States. We used anonymized telemetry data from 40 million (non-business) Windows devices in homes, collected during software updates. These data extend far beyond traditional metrics, such as Internet connectivity or computer ownership, delving into how people interact with technology. This includes activities such as browsing the Internet, using social media, creating documents, and coding software. What emerged from our analysis is that a substantial divide exists between urban and rural communities, not just in infrastructure, but also in our measures of usage and proxies for digital literacy—how effectively and creatively individuals use technology. To capture these differences, we developed two distinct indices: - The Media and Information Composite Index (MCI) measures general computing and media usage, including activities such as browsing, word processing, and emailing. - The Content Creation and Computation Composite Index (CCI) measures specialized activities, including software development and graphic design. Figure 1 demonstrates these geographical disparities, with urban areas generally showing higher levels of usage than rural regions. However, even within major metropolitan areas, such as Chicago (shown in the right panel), substantial variations persist across different ZIP codes, emphasizing that the digital divide is far more severe than infrastructure differences might suggest (broadband availability is much more homogeneous in comparison). <https://pulse-dev-wp.internetsociety.org/wp-content/uploads/2025/07/maps_index.jpg>Figure 1 — Spatial distribution of media and information composite index, content creation and computation composite index, and broadband availability. (A) Media and information composite index across US countries (left) and Chicago-Naperville-Elgin, IL-IN ZIP codes (right). (B) Content creation and computation index across US countries (left) and Chicago-Naperville-Elgin, IL-NL ZIP codes (right). (C) Broadband availability across US counties (left) and Chicago-Naperville-Elgin, IL-IN ZIP codes (right).What Factors Drive These Divides? These findings challenge the assumption that digital divides disappear once basic broadband infrastructure is widely available. Some areas with good Internet infrastructure still experience surprisingly low digital usage. Conversely, regions with robust digital skills can leverage technology for economic and social advancement despite infrastructural limitations. In our analysis, income and education levels emerged as critical influences. Areas with higher household incomes and greater educational attainment consistently showed higher digital engagement on both indices. This suggests that digital usage is closely tied to socioeconomic status, reinforcing existing inequalities. This approach means moving beyond the traditional policy focus on hardware and connectivity. Policymakers, educators, and community organizations must invest in digital literacy programs that cater to the diverse needs of communities. Furthermore, continuous measurement is critical to ensure that initiatives effectively address real-world needs and reassure the audience about their adaptability to changing technological landscapes. This includes continuously monitoring digital usage trends. By making our detailed indices publicly available <https://github.com/microsoft/Consumption-of-Digital-Applications-Data-Set>, we invite further research and policy formulation. The Digital Divide is Evolving Bridging this new divide requires understanding not just who has Internet access, but how they use it. Our findings (read paper <https://ssrn.com/abstract=4957386>) underscore the need for localized strategies that prioritize education, training, and ongoing support to harness technology’s potential for positive change. Contributors: Mayana Pereira, Raffaella Sadun, Tambe Prasanna, Lucia Ronchi Darre, Tammy Glazer, Allen Kim, Rahul Dodhia, and Juan Lavista-Ferres. *Shane Greenstein <https://www.linkedin.com/in/shane-greenstein-30891b6/> is a Martin Marshall Professor of Business Administration at Harvard Business School*. The views expressed by the authors of this blog are their own and do not necessarily reflect the views of the Internet Society. -- -------------------------------------- Joly MacFie +12185659365 -------------------------------------- -