Studies in IT and Economic Development
Paper Session
Friday, Jan. 3, 2025 8:00 AM - 10:00 AM (PST)
- Chair: Diana Moreira, University of California-Davis
Digital Infrastructure and Local Economic Growth: Early Internet in Sub-Saharan Africa
Abstract
We study whether low-speed internet availability fosters local economic growth in rural areas of developing countries by analyzing remote towns in Sub-Saharan Africa. We measure local economic growth of each town by tracking nighttime light emissions. In a difference-in-differences setting, we exploit exogenous countrywide shocks to internet availability induced by submarine cable arrivals in the 2000s and use the rollout of national inter-regional fiber cables to identify towns incidentally connected early. We find that internet availability induces economic growth. Compared to a control group of similar but later connected towns, connected towns experience 11 percent higher light intensity, which translates to 3.3 percentage points higher annual economic growth in the years after internet connection. Additional results suggest this is mainly driven by per-capita productivity growth and not by migration into connected towns. The effect is stronger in towns with better access to regional markets and internet availability is associated with a shift from agriculture to manufacturing in regional employment.E-Reaching the Out of Reach: Evidence on Educating Out of School Children Using Digital Content and Remote Tutoring
Abstract
We estimate the impacts of providing children who have dropped out of school with tablets that have built-in educational software and a phone connection to a private tutor. We conducted a randomized control trial with 890 dropped-out children across Bangladesh. We find that the provision of tablets and tutors increases math scores by approximately 0.25 standard deviations (of the distribution of test scores) and language scores by approximately 0.17 standard deviations. Effects on academic achievement are especially strong for girls, compared to boys, as well as rural children. By contrast, the program did not affect noncognitive traits such as competence, self-esteem, and grit. Taken together, these findings provide one intervention that can improve the educational outcomes of dropped-out children in developing countries, as well as more broadly the educational outcomes of children whose attendance might be affected by temporary school closures due to pandemics or other emergencies.Quantifying the Digital Dividend: The Case of Indian Manufacturing
Abstract
Whether it be in a large-scale automobile plant or a small pottery enterprise, digital technology has become indispensable to business survival. Discussions on digital technology’s economic impact have largely focused on its productivity effects, leading to debates on “Solow’s paradox” and the “productivity miracle” mostly in the context of high-income countries (Stiroh, 2002; Jorgenson, 2011; Acemoglu et al., 2014; Brynjolfsson and McAfee, 2014). However, with the advent of improved and more granular microdata in the past decade, it is now possible to quantify the marginal effects of digital capital in the production process, akin to analyses traditionally performed for labor and physical capital. Using disaggregated data, this study quantifies the digital dividend in the Indian context—an emerging global economy with large formal and informal manufacturing sectors, and a fast-emerging middle-class consumer base. Focusing on digital capital, this study makes several unique contributions. First, the estimated firm-/plant-level technologies incorporate a nested-CES specification for physical and digital capital (allowing for their substitutability) and maintain non-constant returns. Second, a comparison is made leveraging firm and plant-level data from three largest economic databases of India: CMIE (panel), ASI (formal sectors, repeated cross-section) and NSSO (informal sectors, repeated cross-section); covering 2010-2021 for 11 major manufacturing industries and all states. Finally, the simultaneity in inputs is addressed using the control function approach of Levinsohn and Petrin (2003), and Ackerberg, Caves and Frazer (2015). Specification tests are conducted to check the consistency of results.Results point to the presence of a robust, statistically and economically significant digital dividend, with the estimated elasticity ranging from 0.02 to 0.17 for industries in the formal sector, and 0.02 to 0.05 for industries in the informal sector. Moreover, results extend the self-selection hypothesis—firms with higher productivity (than the industry average) tend to be large exporters— to the case of digital capital.
The Economic Consequences of IT: The IT Revolution’s Meager Benefits and Major Schisms
Abstract
The IT revolution, underway since around 1980, has featured mediocre growth and rising geographic, educational, and generational inequality. This stands in stark contrast to the broad prosperity and convergence experienced in the 1950s and 1960s. We attribute this change to a swivel in the leading edge of productivity growth away from manufacturing largely present in mid-sized towns to information technology mainly housed in “superstar” cities. Using a spatial model, we show how this can explain: rising prosperity and rapid housing inflation in superstar cities; falling relative wages in towns and the countryside; mediocre aggregate productivity due to increasing misallocation of labor; increasing labor market division as skilled workers crowd into cities and unskilled workers move out; and falling internal migration. Empirical evidence on the geographic evolution of US wages, house prices, technological change, and internal migration strongly support the predictions of the model.JEL Classifications
- O1 - Economic Development