This dissertation examines the development outcomes of mobile internet following its fast diffusion in low-middle-income countries (LMICs) in the early 2010s – early 2020s. In developing countries, many low-income and rural populations have skipped laptops/desktops and adopted smartphones as their first means of engaging in the digital world. The dissertation comprises of three main studies. The first study provides evidence of the positive effects on average household income per capita in Viet Nam, using province and time-fixed effects combined with instrumental variable (FE-IV) to alleviate endogeneity concerns. Heterogeneity analysis demonstrates the more inclusive impacts — stronger effect sizes for rural areas and lower-income quintile groups. Key mechanisms detected include improvements in employment, particularly skilled jobs and wages, as drivers of household income effects. The chapter adds to the second-generation mobile for development (M4D 2.0) literature both rigorous empirical evidence and economy-wide impacts at province or regional level. The second study is possibly among the first analyses to examine the educational attainment outcomes of 3G internet coverage. Using FE-IV estimation with lightning strikes per capita as an instrument, we find that 3G diffusion in Viet Nam discourages uppersecondary education while encouraging lower-secondary attainment. Youth aged 15-25 who completed lower-secondary school opted to enter the workforce rather than pursue or complete upper-secondary education, driven by new job opportunities and higher wages from 3G technology diffusion. This underscores the opportunity cost of education and highlights the multidimensional effects of mobile internet diffusion in LMICs. The third study sheds light on the negative effects of internet access on youth mental health and subjective well-being in Ethiopia. Heterogeneity analysis reveals unequal impacts, with adolescents from lower-wealth households experiencing worsened negative outcomes. ‘Mindless scrolling’ or passive internet use are detected as an important mechanism causing such unequal effects. Using FE-IV with community’s average internet adoption rates as IV, we alleviate endogeneity concerns related to unobserved heterogeneities and mismeasurement bias. As far as we know, this is among the first papers offering robust evidence of internet use’s impact on youth’s mental health in a low-income country context.
Overall, the dissertation underlines both the pro-poor economic outcomes of mobile internet diffusion on income per capita from the (regional) economywide perspective, and the negative impacts on youth’s education and mental health, with unequal effects particularly among youth from less advantaged socioeconomic backgrounds. The findings illuminate the subtle narrative of internet as it evolves towards universal coverage, thereby necessitating comprehensive assessments across different demographics and geographies. The results have significant implications for national and regional policies regarding the socioeconomic impacts of technology diffusion, particularly of
a general-purpose technology (GPT) like mobile internet. Such analyses are essential for fostering inclusive development and inclusive innovation that goes beyond simply connectivity, ensuring that the benefits of technological advancements are widely and more equitably distributed.
