Page 175 - ISC PROCEEDINGS 21.4

INTERNET PENETRATION AND INCOME INEQUALITY: EVIDENCE OF A
TURNING POINT FROM GLOBAL PANEL DATA

4
3
1
Duong Duc Anh* , Le Thi Ngoan , Ngo Thi Cam Van , Phan Thi Huong ,
2
Tran Thi Khanh Tung 5
1, 3, 4, 5 Vinh University, Vinh, Vietnam.
2 Kobe University, Kobe, Japan.
(*E-mail: anhdd@vinhuni.edu.vn)

ABSTRACT
This paper investigates the relationship between internet penetration and income
inequality using an unbalanced panel of 128 countries over 2000–2022. Estimating a
quadratic two-way fixed effects model with Driscoll–Kraay (1998) standard errors robust
to cross-sectional dependence, we document a statistically significant U-shaped
relationship — consistent with a Digital Kuznets Curve (DKC) — between internet
penetration and the Gini coefficient, with a global turning point at approximately 62
percent of the population (z = 2.55, p = 0.011). An income-group interaction model
reveals that this threshold is reached earlier in high-income countries (≈52%) than in
middle-income economies (≈66%), consistent with the mediating role of institutional
quality. Among alternative ICT measures, fixed broadband exhibits a nonlinear pattern
with a lower turning point (≈13 subscriptions per 100), whereas mobile cellular
subscriptions display no significant nonlinearity — identifying cost barriers as the
mechanism driving technology-specific distributional dynamics. System GMM estimates
are consistent with digital diffusion affecting inequality through gradual, long-run
channels.
Keywords: Digital Kuznets Curve; Gini coefficient; income inequality; internet
penetration; panel data.

1. Introduction
Income inequality remains one of the most consequential issues in contemporary
economic policy. Despite sustained economic growth across much of the developing
world, the World Bank's Poverty and Inequality Platform records that within-country
inequality has widened in a substantial number of economies (World Bank, 2024). The
share of the global population using the internet rose from under 10 percent in 2000 to
approximately 65 percent by 2022 — a transformation that has fundamentally altered
how individuals access information, participate in labour markets, and engage with
financial and public services. Whether this transformation has reduced or exacerbated
income inequality remains an open empirical question (Njangang et al., 2022; Ho et al.,
2025).
The theoretical priors are ambiguous. Internet access reduces information
asymmetries and expands market participation, suggesting an equalising effect (Hjort &
Poulsen, 2019). Conversely, the skill-biased technological change (SBTC) hypothesis
predicts that digital technologies primarily augment skilled workers' productivity,
widening income differentials (Acemoglu, 2002). Extending Kuznets (1955) to digital
diffusion, this paper hypothesises a U-shaped Digital Kuznets Curve (DKC): internet
penetration initially reduces the Gini coefficient (β₁ < 0), but as coverage reaches high

174