Abstract

Despite rapid expansion of digital infrastructure, meaningful access to information and communication technologies (ICTs) remains uneven across population groups. Digital inclusion is not only a question of network availability, but also of affordability, skills, device access, and the social and economic contexts in which individuals and households live. This paper presents a human-centred analysis of the digital divide using nationally representative household survey data, with the aim of translating ICT statistics into actionable evidence for inclusive digital policy.

The analysis draws on the 2023/24 Kenya Housing Survey (KHS), a large-scale, nationally representative survey covering all 47 counties and over 25,000 households. The ICT module of the survey captures ownership and use of mobile phones, computers, and the Internet at both individual and household levels, alongside information on education, age, sex, disability status, employment, housing characteristics, and household living conditions. This rich socio-economic context allows ICT outcomes to be interpreted not merely as technical indicators, but as reflections of people’s lived realities.

We focus on three interrelated dimensions of digital inclusion. First, we examine who remains offline or under-connected, using standard indicators of internet use, computer use, and household internet access, disaggregated by residence, education level, age group, income quintile proxies, disability status, and housing characteristics. The results reveal persistent and intersecting inequalities, with markedly lower ICT use among rural populations, individuals with lower education attainment, older adults, persons with disabilities, and residents of informal or traditional housing structures. Second, we assess device access and adequacy, highlighting the gap between basic mobile phone ownership and access to devices that enable effective participation in digital services, education, and employment. Third, we analyse reported barriers to internet use, grouping them into policy-relevant categories such as affordability, coverage and network quality, device constraints, and skills or awareness limitations.

To move beyond descriptive statistics, the paper examines how different socio-economic factors contribute to observed gaps in internet use and device access. By comparing outcomes across population subgroups, we identify which characteristics are most strongly associated with digital exclusion and how these associations vary across contexts. This approach allows us to distinguish between disparities driven by structural living conditions (such as location and housing type) and those linked to individual characteristics (such as education or age), providing a clearer basis for policy prioritisation.

The paper concludes by linking statistical findings to concrete intervention pathways. Rather than a one-size-fits-all approach, the results point to differentiated strategies: affordability-focused interventions for low-income households, last-mile connectivity improvements in rural and remote areas, device access initiatives where hardware constraints dominate, and digital skills programmes tailored to older populations and marginalised groups. By grounding ICT indicators in human and household contexts, this paper demonstrates how official statistics can more effectively inform inclusive digital transformation, ensuring that progress in connectivity translates into improved opportunities and quality of life for all.