What do we know?
To date, much of the existing literature has focussed on the potential of internet technology to ‘jump-start’ the development process in low- and middle-income countries (LMICs). A number of research studies have evaluated the impact of access to fixed high-speed broadband internet on socio-economic outcomes. However, understanding the impact of mobile internet, which is far more accessible and far more prevalent than broadband internet in LMICs, has been limited.
Early studies assessing the impact of better 3G coverage focussed more on its impact on political mobilisation (Zhuravshkaya et al. 2020, Guriev et al. 2021). Recent studies assessing the economic impacts of this technology have focussed on a few individual countries (mostly in Africa) and, for the most part, on outcomes other than employment (Bahia et al. 2020, Masaki et al. 2020, Mensah 2021, Viollaz et al. 2022).
In our study, we attempt to address these knowledge gaps. We aim to evaluate the causal effects of 3G internet on employment outcomes of individuals across LMICs. We are particularly interested in assessing whether 3G access promotes structural transformation in the sense of shifting employment away from agriculture and towards manufacturing and services, and investigating how such a transition might affect women.
Our study uses a sample of 14 countries in various stages of economic development, ranging from low-income countries such as Zambia and Uganda; lower-middle-income countries such as Indonesia, Bolivia, and the Philippines; and upper-middle income countries, such as Brazil and Mexico. Across these countries, our sample covers over a billion people, including 600 million working-age individuals (that is, from ages 18-65). We divide our sample into smaller sub-national regions such as districts, counties and municipalities. After restricting the sample to those sub-regions where all variables can be observed (2G and 3G coverage, employment data, etc.), our final sample consists of 6,802 regions and 16,069 region-years evaluated from the period 2000-2015.
Challenges in identifying a causal relationship
Drawing a causal relationship between access to 3G internet and employment outcomes presents three key challenges.
First, there is a lack of reliable data across multiple low-income countries and especially over a longer period of time. We address this by utilising data from IPUMS International, which collates nationally representative surveys and censuses across multiple countries and over time. This allows us to construct key employment outcomes relevant to our study (such as labour force participation rates, types of employment, etc.) that are both harmonised across all countries in our sample, and consistent over time.
Second, there is little available data on the expansion of 3G coverage, especially at a local or sub-national level. To address this issue, we use maps for 3G network coverage from 2006-2015 collected by Collins Bartholomew Mobile Coverage Explorer, which consist of one square kilometre binary grids that show whether the cell region has 3G coverage or not. We then aggregate these to generate a (population-weighted) measure of 3G coverage for each sub-national region over time.
Lastly, it can be difficult to identify a causal relationship between 3G internet and employment due to the endogenous expansion of 3G networks with economic development. For example, it is possible that regions with higher economic activity are also the ones that are more likely to get access to 3G internet. This is known as ‘reverse causality’. A common method used to address this issue is an instrumental variable (IV) strategy1. Following the literature (Manacorda et al. 2020, Guriev et al. 2021), we use lightning strikes as an instrument – we argue that conditional on geographic factors (such as elevation, precipitation, etc.) the intensity of lightning strikes in an area affects the rollout of 3G network, but does not affect employment outcomes directly.2 In addition, we also control for 2G coverage in our specifications, so as to isolate the effects arising specifically from the availability of 3G, as opposed to other factors that may have affected cellphone expansion more generally.
In sum, a combination of harmonised, rich data across multiple countries and sub-national regions, a novel dataset that measures the expansion of 3G networks across these regions, and the IV strategy together allow us to causally examine the relationship between 3G access and a key set of employment outcomes, both for men and women.
Nuances of the gendered employment effects of 3G access
Overall, we find that 3G coverage has had beneficial effects on employment outcomes and opportunities for both men and women. Increased 3G coverage has demonstrably increased the employment rates of both genders, and meaningfully increased female labour force participation rates.
Our results also indicate that increased 3G coverage has affected the nature of employment, but as expected and hypothesised by many researchers and policymakers, we do not find any evidence of ‘structural transformation’ (that is, the reallocation of labour away from agriculture, and towards manufacturing and services). If anything, 3G expansion has created additional jobs in both agriculture and services.
The gender dimension of these changes play out in two ways. When men transition out of unpaid agricultural and service jobs towards self-employment (small owner-operated businesses) in agriculture or wage jobs in the service sector, women are more likely to take up the unpaid agricultural jobs vacated by men. However, women also start small enterprises in agriculture and move into wage employment in the service industry – the magnitude of the effects are comparable to each other.
Put together, these results indicate that rather than jump-starting structural transformation that would affect the sector of employment, 3G coverage impacts the type of employment (whether it is unpaid jobs, wage jobs, or self-employment), and does so differently for men and women.
Looking forward: Gender and the future of work
The gendered effect of increased 3G coverage is notable. It may be that increased 3G internet connectivity allows women to balance household responsibilities and work better, thus allowing them to enter the labour force and operate small businesses, perhaps remotely from home. These patterns are also consistent with the emergence of a gig-economy, with more flexible work opportunities that could be valued more by women. On the other hand, women taking up the unpaid jobs vacated by men raises questions regarding the short- and long-term transitions in the labour market and the resulting gender inequalities arising in LMICs as a result.
To conclude, our study presents some intriguing patterns on the impact of an important technology that has disrupted labour markets across the world. These patterns raise several important questions about the mechanisms underlying the documented changes, which could lead to an exciting research agenda in the future to help us understand how mobile internet is changing the future of work, particularly for women.
The authors thank Vanika Mahesh, Communications Intern at the Yale Economic Growth Center, for significant editorial assistance.
Instrumental variables are used to address endogeneity concerns by generating a plausible source of exogenous variation. An instrument is correlated with the explanatory factor – in this case, the rollout of 3G internet – but does not directly affect the outcome of interest, and thus can be used to estimate the causal relationship between the two.The equipment needed for mobile phone infrastructure is particularly sensitive to electrical surges due to frequent lightning strikes, and is expensive to repair if damaged. Moreover, lightning severely interferes with the quality of radio wave transmissions, which affects the quality of reception. These factors make it less likely that 3G networks would be installed in regions with a high frequency of lightning strikes
- Goldberg, PK and G Chiplunkar (2022), ‘The Employment Effects of Mobile Internet in Developing Countries’, NBER Working Paper 30741.
- Bahia, K, P Castells, G Cruz, X Pedros, T Pfutze, C Rodríguez-Castelán and H Winkler (2020), ‘The Welfare Effects of Mobile Broadband Internet: Evidence from Nigeria’, World Bank Policy Research Working Paper 9230.
- Guriev, Sergei, Nikita Melnikov and Ekaterina Zhuravskaya (2021), "3G Internet and Confidence in Government", The Quarterly Journal of Economics,136(4): 2533-2613. Available here.
- Manacorda, Marco and Andrea Tesei (2020), “Liberation Technology: Mobile phones and Political Mobilization in Africa”, Econometrica, 88(2): 533–567.
- Masaki, T, R Ochoa and C Rodríguez-Castelán (2020), ‘Broadband Internet and Household Welfare in Senegal’, World Bank Policy Research Working Paper 9386.
- Mensah, JT (2021), ‘Mobile Phones and Local Economic Development: A Global Evidence’, SSRN 3811765.
- Zhuravskaya, Ekaterina, Maria Petrova and Ruben Enikolopov (2020), "Political Effects of the Internet and Social Media”, Annual Review of Economics, 12: 415-438.
- Viollaz, Mariana and Hernan Winkler (2022), "Does the Internet Reduce Gender Gaps? The Case of Jordan”, The Journal of Development Studies, 58(3): 436-453.