Development Informatics

Editorial commentary on ICT for development research. Not affiliated with the former IDIA conference series.

Women's Digital Inclusion After the Baku Roadmap: What the Gender Gap in Mobile Internet Use Really Measures

Two abstract bar charts side by side in teal and orange representing a gap between two groups

Two respected sources describe the gender gap in internet use, and their headline numbers do not match. The International Telecommunication Union (ITU) reported in its Facts and Figures 2025 that 77% of men and 71% of women worldwide use the Internet, a difference of about 280 million people. The GSMA’s Mobile Gender Gap Report 2025 found that women in low- and middle-income countries are 14% less likely than men to use mobile internet, which it puts at 235 million fewer women than men.

Neither is wrong. They measure different things, in different populations, with different methods. After the ITU’s World Telecommunication Development Conference in Baku adopted its 2026 to 2029 Action Plan, the distinction is worth understanding, because the plan will be judged against indicators like these.

Two different yardsticks

The ITU’s measure is the gender parity score: the percentage of women using the Internet divided by the percentage of men. Parity is deemed achieved when the score falls between 0.98 and 1.02. The global score in 2025 was 0.92, the same as in 2019, which the ITU reads as no global progress toward parity in six years. Regionally, the picture is uneven. Gender parity has been achieved in Europe, the Americas and the CIS region, while the score for Africa rose from 0.70 to 0.78 over the period and the Asia-Pacific score slipped from 0.92 to 0.91.

The GSMA’s measure is a relative gap in mobile internet use among adults in low- and middle-income countries, drawn from its own survey programme rather than from national statistical offices. By its figures, 63% of women in those countries now use mobile internet, more than ever before, yet the pace of adoption slowed in the past year. The gap narrowed from 25% to 15% between 2017 and 2020, then stalled and widened slightly in 2021 and 2022.

The differences are structural. The ITU reports on Internet use by any means, across all countries, from official data. The GSMA reports on mobile internet in a defined group of lower-income markets. A smaller gap in one and a larger gap in the other is what you would expect when one series includes high-income countries at parity and the other does not. Comparing the two as if they were competing estimates of the same quantity is a category error.

What the gap does not measure

Both indicators count whether someone uses the Internet, which is a threshold question. They say little about the quality of use.

The GSMA report itself points past the threshold. Even once women are online, it notes, they tend to use mobile internet less frequently than men and for a narrower range of services, citing safety and security concerns, affordability and connectivity experience. Handset economics are one concrete driver: the report finds that an entry-level handset costs 24% of a woman’s monthly income in these markets against 12% of a man’s. Smartphone ownership shows its own gap of 14%, with 61% of women owning a smartphone and around 945 million women without one.

A binary “uses / does not use” figure therefore hides a gradient: shared devices, intermittent data, limited skills and constrained autonomy over what one does online. Qualitative studies have long argued that simple ownership or usage statistics miss these experiences. The Facts and Figures and the GSMA survey are both necessary, and neither is sufficient for judging whether women’s digital inclusion is meaningful.

What Baku changes, and what it does not

The Baku Declaration and Action Plan set the ITU’s development-sector priorities for 2026 through 2029. The conference included a dedicated session of the Network of Women in ITU-D, and the plan’s regional initiatives cover themes that have recurred across cycles, including gender-inclusive connectivity. The plan frames its goal as universal, meaningful and affordable connectivity, but it does not attach an end date to that goal.

In the conference summaries available at the time of writing, we found no dated, numeric gender target. Readers should check the full text of the resolutions and regional initiatives, which are the operational parts of the package, before assuming one exists. If none does, the practical consequence is that progress will be tracked through whichever indicators institutions choose to report, and the choice between the ITU’s parity score and a survey-based mobile gap will shape the narrative.

Reading the numbers in practice

For a researcher or program officer, a few working rules follow from the above.

Name the population. “Worldwide,” “low- and middle-income countries” and “least developed countries” produce different gaps. The ITU notes that least developed and landlocked developing countries show only small improvement, with men still significantly more likely than women to use the Internet.

Name the instrument. Official household surveys, operator administrative data and commercial survey programmes differ in coverage and definitions. State which one a claim rests on.

Separate access from use. A woman who owns a phone, one who can use it for data and one who does so safely and regularly are three different people in the statistics.

Treat stalls as findings. An unchanged parity score across six years, during which total user numbers rose by roughly 45% for both men and women, says that growth alone does not close the gap. Expansion is lifting both groups at similar rates.

The takeaway for evaluation design is to pair a headline gap with measures of depth: frequency of use, range of services, device ownership and reported safety. A single parity ratio will not tell a funder whether an intervention changed anything for women. The published ITU analysis of the gender digital divide and the GSMA Mobile Gender Gap Report are the two primary references, and a careful study will cite both and explain why they differ.