The research, conducted by Thi Thu Trang Dao, examined a candidate panel covering 131 developing economies during the period 2005 to 2021. The fully adjusted model drew on 1291 country-year observations from 96 economies, allowing the analysis to test whether rising Internet use is associated with changes in ANS relative to gross national income. The core estimate came from two-way fixed-effects regressions with country-clustered inference.

In the fully adjusted model, a one-percentage-point increase in lagged Internet use was associated with a change of −0.0539 percentage points in ANS/GNI, with a 95 percent confidence interval running from −0.1336 to 0.0259. Because that interval spans zero, the average association is not statistically distinguishable from no effect at conventional thresholds. The author did not stop at a single headline number.

The estimate proved stable when individual countries were excluded one at a time and under alternative income classifications, suggesting the result is not driven by a handful of outliers or by how economies are sorted into income groups. However, when country-specific trend specifications were applied, the sign of the association flipped, indicating that the apparent average relationship depends heavily on how underlying time trends are modeled. To probe whether the association differed across contexts, the study ran common-sample interaction tests with multiplicity correction.

No primary moderator survived either Benjamini–Hochberg or Holm correction, meaning none of the tested conditioning factors provided reliable evidence that Internet adoption's association with adjusted net savings varies systematically across subgroups. One of the study's most revealing contributions is an exact common-sample decomposition of the ANS measure into its constituent parts. This analysis showed offsetting associations beneath the aggregate figure.

The association involving particulate-emission damage remained statistically detectable after both correction procedures, while the association involving forest depletion survived only the false discovery rate correction and not the stricter Holm procedure. These decomposition results suggest that digital access may be associated with different components of national wealth accounting in opposing directions, with some channels offsetting others and leaving the net effect close to zero on average. The study frames this as evidence that the sustainability implications of digital access are multidimensional and sensitive to trend treatment rather than captured by one uniform directional effect.

For researchers and policymakers monitoring sustainable development goals, the findings argue for caution in interpreting digital connectivity as an unambiguous driver of sustainable wealth accumulation.

Subject of Research: The relationship between Internet adoption and adjusted net savings in developing economies

Article Title: Digital connectivity and adjusted net savings in developing economies

Article References: Dao, T. T. T. (2026). Digital connectivity and adjusted net savings in developing economies. Discover Sustainability. https://doi.org/10.1007/s43621-026-05135-z

Image Credits: AI Generated

DOI: 10.1007/s43621-026-05135-z

Keywords: digital connectivity, adjusted net savings, developing economies, Internet adoption, wealth accounting, sustainability, panel analysis, fixed effects, SDG monitoring, pollution damages, Digital, connectivity