Crisis communication and internet penetration: Cross-country evidence with illustrations from Moldova and Armenia

  • 9 Views
  • 0 Downloads

Creative Commons License DMCA.com Protection Status
This work is licensed under a Creative Commons Attribution 4.0 International License

Type of the article: Research Article

Abstract
Coordinated public information campaigns became a near-universal instrument of pandemic crisis communication, while the internet penetration of their audiences ranged from under 3 percent to full coverage. The study examines whether the association between campaign intensity and the pace of vaccination uptake varies with pre-pandemic internet penetration, a contextual marker of the digital environment rather than campaign exposure. Using a global country-month panel of 168 economies over 2021–2022 (N = 3,410), it estimates two-way fixed-effects models with country-clustered standard errors, drawing campaign intensity from the Oxford COVID-19 Government Response Tracker and fixing connectivity at its 2019 level. The interaction between campaigns and internet penetration is positive and robust (0.912, p < 0.001): a standard deviation of baseline penetration, 27.9 percentage points, raises the campaign–uptake slope by about 0.9 points per month. The marginal association is significantly negative at low connectivity (−1.955 at the tenth percentile, p < 0.001), indistinguishable from zero at the median (−0.289) and the ninetieth percentile (+0.501), with the confidence band excluding zero only below roughly 62 percent penetration; the firm result is the negative association under low connectivity. The pattern survives winsorizing, one-month lagging, separate estimation for 2021 and 2022, and vaccination-policy controls, yet baseline connectivity correlates with income and governance at r = 0.77–0.90, so the moderation indexes a development bundle of which connectivity is one legible marker. The illustrative cases of Moldova and Armenia, comparably connected yet facing contrasting overlapping crises, show connectivity alone does not assure uptake where parallel emergencies compete for attention.

Acknowledgments
The authors acknowledge the financial support received for the implementation of the project “Accelerating the Digital and Green Transformation of Agri-Food SMEs in Moldova and Armenia”, funded by the National Agency for Research and Development of the Republic of Moldova (NARD) (Project No. 26.80013.0807.05ARM) and the Higher Education and Science Committee of the Ministry of Education, Science, Culture and Sports of the Republic of Armenia (Project No. 26NARD-1D007).

view full abstract hide full abstract
    • Figure 1. Marginal effect of public information campaigns (H1) on vaccination pace across pre-pandemic (2019) internet penetration, with 95% confidence interval (column 2 of Table 3)
    • Figure 2. Moldova: COVID-19 crisis communication intensity (public information campaigns and containment stringency) and vaccination dynamics, 2020–2022
    • Figure 3. Digital channel capacity, Moldova versus Armenia: e-government sub-indices and internet penetration, 2024
    • Figure 4. Armenia, overlapping crises: monthly political-violence events (upper panel) and Nagorno-Karabakh conflict fatalities (lower panel), 2019–2024, with the COVID-19 window shaded
    • Figure 5. Moldova’s 2022 economic shock: consumer-price inflation (upper panel) and real GDP growth (lower panel), Moldova versus Armenia, 2015–2024
    • Figure C1. Model-implied marginal effect of campaigns for a representative spread of economies across the 2019 connectivity distribution (23 economies, every eighth by connectivity rank; Moldova highlighted). Point estimates with 95% confidence intervals
    • Figure D1. Leave-one-country-out estimates of the campaign × connectivity interaction (sorted; 95% CI; M2)
    • Table 1. Variables, definitions and sources
    • Table 2. Descriptive statistics (analysis sample, 2021–2022)
    • Table 3. Fixed-effects estimates of the campaign–vaccination relationship
    • Table 4. Marginal effect of public information campaigns (H1) on vaccination pace across the 2019 internet-penetration distribution
    • Table 5. Robustness and scope checks (coefficient on the H1 × internet interaction)
    • Table A1. Economies in the estimation sample, with income group, pre-pandemic (2019) internet penetration, and connectivity group
    • Table B1. Pearson correlations among the model variables (estimation sample, N = 3,410)
    • Table C1. Model-implied marginal effect of campaigns, all economies
    • Table D1. Leave-one-country-out interaction coefficient, all economies (alphabetical)
    • Conceptualization
      Liliana Staver, Tatul M. Mkrtchyan, Larisa Dodu-Gugea, Artiom Jucov, Mihaela Pascal, Anton Lia, Elena Ciochina
    • Formal Analysis
      Liliana Staver, Elena Ciochina
    • Funding acquisition
      Liliana Staver, Tatul M. Mkrtchyan
    • Supervision
      Liliana Staver
    • Writing – original draft
      Liliana Staver, Tatul M. Mkrtchyan, Larisa Dodu-Gugea, Artiom Jucov, Mihaela Pascal, Anton Lia, Elena Ciochina
    • Writing – review & editing
      Liliana Staver, Tatul M. Mkrtchyan, Larisa Dodu-Gugea, Artiom Jucov, Mihaela Pascal, Anton Lia, Elena Ciochina
    • Data curation
      Tatul M. Mkrtchyan, Anton Lia
    • Methodology
      Tatul M. Mkrtchyan
    • Project administration
      Tatul M. Mkrtchyan
    • Validation
      Larisa Dodu-Gugea, Mihaela Pascal
    • Software
      Artiom Jucov, Elena Ciochina
    • Visualization
      Artiom Jucov, Anton Lia
    • Resources
      Mihaela Pascal