Publications
Roads, Competition, and the Informal Sector
Perra E., Sanfilippo, M. Sundaram, A. (2024) "Roads, Competition, and the Informal Sector", Journal of Development Economics
We examine the impact of competition from better connectivity to domestic markets on formal and informal firms. Combining geolocalized information on road improvements under a large infrastructure investment programme with data on manufacturing firms in Ethiopia between 2001 and 2013, we show that an increase in competition is associated with higher labour productivity, capital-intensity, investment in physical capital and wages in the formal sector. On the contrary, there is no associated increase in labour productivity or wages in the informal sector. In fact, increased competition results in lower capital-intensity and investment, a shift in composition towards workers without primary education and a lower likelihood of operating in the informal sector. We thus highlight that the benefits of infrastructure improvement programmes may not accrue uniformly in the economy.
We examine the impact of competition from better connectivity to domestic markets on formal and informal firms. Combining geolocalized information on road improvements under a large infrastructure investment programme with data on manufacturing firms in Ethiopia between 2001 and 2013, we show that an increase in competition is associated with higher labour productivity, capital-intensity, investment in physical capital and wages in the formal sector. On the contrary, there is no associated increase in labour productivity or wages in the informal sector. In fact, increased competition results in lower capital-intensity and investment, a shift in composition towards workers without primary education and a lower likelihood of operating in the informal sector. We thus highlight that the benefits of infrastructure improvement programmes may not accrue uniformly in the economy.
Road to division: Ethnic favoritism and road infrastructure in Ethiopia
Perra E. (2026) "Road to division: Ethnic favoritism and road infrastructure in Ethiopia", Journal of Comparative Economics
This paper investigates the role of ethnic favoritism in the long run allocation of road infrastructure in Ethiopia. I construct a 5 km grid cell panel by merging road network data from the late 1960s to 2016 with high resolution maps of local ethnic composition. Using a quasi-experimental empirical design, the study finds that cells where the local majority shares the ethnicity with the ruling elite receive 8.7 percent additional road investments and see a 13.8 improvement in road surface quality than otherwise comparable non co-ethnic cells. Exploiting the phased rollout of the Road Sector Development Program between 1997 and 2016, I also document that new roads raise night lights intensity by 0.27 standard deviations in cells benefiting from the program and by 0.51 standard deviations in co-ethnic areas, with larger effects for earlier investments. These findings suggest that ethnic alignment and the availability of large public funds jointly shape the placement of road infrastructure, influencing local economic dynamics.
This paper investigates the role of ethnic favoritism in the long run allocation of road infrastructure in Ethiopia. I construct a 5 km grid cell panel by merging road network data from the late 1960s to 2016 with high resolution maps of local ethnic composition. Using a quasi-experimental empirical design, the study finds that cells where the local majority shares the ethnicity with the ruling elite receive 8.7 percent additional road investments and see a 13.8 improvement in road surface quality than otherwise comparable non co-ethnic cells. Exploiting the phased rollout of the Road Sector Development Program between 1997 and 2016, I also document that new roads raise night lights intensity by 0.27 standard deviations in cells benefiting from the program and by 0.51 standard deviations in co-ethnic areas, with larger effects for earlier investments. These findings suggest that ethnic alignment and the availability of large public funds jointly shape the placement of road infrastructure, influencing local economic dynamics.
Research Pipeline
Beyond Data Gaps: Tracking Spatial Inequality in Africa via Nighttime Lights
Martorano, B., Perra, E., and Tiberti, M., (2023) "Spatial inequality during the COVID-19 pandemic in Africa using night-time lights data", MERIT Working Papers 2023-032, United Nations University.
This paper offers a robust approach to timely measure the distributional impact of crises at a sub-national level, offering a valuable alternative to traditional economic measures that are often delayed or incomplete. Using the COVID-19 pandemic as an illustrative case study, we first employ satellite-derived nighttime light data to measure real-time economic disparities at the local level across African countries. We show that national measures of inequality mask substantial heterogeneity at sub-national level. We then combine this information with the novel Causal-ARIMA (Causal-AutoRegressive Integrated Moving Average) estimator to assess if there is any relationship between the outbreak of the health crisis, the strictness of policy restrictions and the changes observed in spatial inequality. The results indicate that wealthier and more industrialized areas implementing more stringent containment measures experienced more pronounced reductions in inequality. The study underscores the potential of our empirical strategy for tracking inequality dynamics during times of crisis, offering a valuable tool for policymakers when canonical data sources are inaccessible.
This paper offers a robust approach to timely measure the distributional impact of crises at a sub-national level, offering a valuable alternative to traditional economic measures that are often delayed or incomplete. Using the COVID-19 pandemic as an illustrative case study, we first employ satellite-derived nighttime light data to measure real-time economic disparities at the local level across African countries. We show that national measures of inequality mask substantial heterogeneity at sub-national level. We then combine this information with the novel Causal-ARIMA (Causal-AutoRegressive Integrated Moving Average) estimator to assess if there is any relationship between the outbreak of the health crisis, the strictness of policy restrictions and the changes observed in spatial inequality. The results indicate that wealthier and more industrialized areas implementing more stringent containment measures experienced more pronounced reductions in inequality. The study underscores the potential of our empirical strategy for tracking inequality dynamics during times of crisis, offering a valuable tool for policymakers when canonical data sources are inaccessible.
Work in Progress
The Fertility Highway: How Road Investments Shape Childbearing in Ethiopia
Hiluf Abay, M., Perra, E. and Tiberti, L. (2026) "The Fertility Highway: How Road Investments Shape Childbearing in Ethiopia"
This paper examines the impact of road infrastructure expansion on fertility decisions in Ethiopia. Using geo-referenced road network data combined with the Demographic and Health Surveys (DHS), we investigate fertility outcomes of road improvements under the Road Sector Development Program (RSDP). Employing instrumental variable and dynamic difference-in-differences methods, we find that road improvements significantly reduce fertility rates. A 10 kilometre increase in road length lowers completed fertility by 0.54 children and decreases the annual probability of childbirth by 1.3 percentage points, corresponding to a 12% decrease. Roads also improve child survival, delay women’s first marriage and childbirth, and boost female employment and schooling.
This paper examines the impact of road infrastructure expansion on fertility decisions in Ethiopia. Using geo-referenced road network data combined with the Demographic and Health Surveys (DHS), we investigate fertility outcomes of road improvements under the Road Sector Development Program (RSDP). Employing instrumental variable and dynamic difference-in-differences methods, we find that road improvements significantly reduce fertility rates. A 10 kilometre increase in road length lowers completed fertility by 0.54 children and decreases the annual probability of childbirth by 1.3 percentage points, corresponding to a 12% decrease. Roads also improve child survival, delay women’s first marriage and childbirth, and boost female employment and schooling.
Oil Spills, Water Networks and Local Economic Development
Sanfilippo, M., Perra, E. and Sileci, L. (2026) "Oil Spills, Water Networks and Local Economic Development"