This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2027
  4. Preliminary Program
  5. Agricultural Productivity and Structural Transformation in Developing Countries
« Back to Results

Agricultural Productivity and Structural Transformation in Developing Countries

Paper Session

Sunday, Jan. 3, 2027 8:00 AM - 10:00 AM (EST)

Grand Hyatt Washington
Hosted By: Agricultural and Applied Economics Association
  • Chair: Jeremy Foltz, University of Wisconsin-Madison

Gendered Constraints and Structural Transformation: Evidence from Agricultural Productivity in Sub-Saharan Africa

Chunxiao Jing
,
University of Wisconsin-Madison

Abstract

Structural transformation—labor reallocation from low-productivity agriculture to more productive manufacturing and services—is a central driver of economic growth in low- and middle-income countries. While recent work (Buera et al., 2011; Buera & Shin, 2013) highlights household-level constraints that can limit labor reallocation, these constraints may be particularly binding for women. Female participation in non-agricultural sectors remains low and has stagnated across much of Sub-Saharan Africa (Backhaus & Loichinger, 2022; Asaleye & Strydom, 2024). However, it remains unclear whether these gender differences reflect preferences or binding constraints. This paper asks: Do gender differences in labor allocation reflect preferences or constraints, and to what extent do gender-specific constraints limit women’s labor reallocation during structural transformation?
We construct individual panel data for four African countries (Niger, Nigeria, Malawi, and Ethiopia) using the World Bank’s LSMS surveys, combined with high-resolution satellite data on NDVI (MODIS), rainfall (CHIRPS), and temperature (ERA-LAND). We develop an exogenous and continuous measure of weather-driven agricultural productivity by predicting NDVI growth from rainfall and temperature using a Random Forest model. We then model sectoral transitions as the outcome of a household decision problem in which men and women face identical wages but gender-specific entry costs into non-agricultural sectors. By interacting gender with the productivity measure, we identify heterogeneity in these entry costs.
We find that women respond more strongly than men to increases in weather-induced agricultural productivity by transitioning into non-agricultural sectors. This asymmetric response suggests that women face binding entry constraints in low-productivity states that are partially relaxed when agricultural productivity improves. The results are inconsistent with purely preference-driven differences and instead point to gender-specific barriers to mobility. These findings imply that such constraints can significantly slow structural transformation by limiting women’s movement into more productive sectors.

Do Floods Induce Long-term Livelihood Diversification? Evidence from Nigeria

Pierre Biscaye
,
Université Clermont Auvergne
Abdulrasheed Isah
,
AAEA

Abstract

Floods are among the most damaging climate-related shocks in low- and middle-income countries, but it remains unclear whether severe flood exposure induces households to diversify away from agriculture over the long term. This paper examines how flooding affects household livelihoods, labor allocation, and consumption in Nigeria using a nationally representative household panel spanning 2010–2019. We estimate impacts in a difference-in-differences framework that controls for predicted flood exposure and compare community-level flooding measured using survey reports of harmful flooding and a widely used database of satellite-detected inundation. Using survey-based flood exposure, we find persistent declines in consumption (12%) and agricultural production (17%) with no increase in non-farm enterprise labor or earnings. Wage employment rises only among households with prior wage work and without corresponding increases in wage earnings, while migration for work only increases after a delay of several years. In contrast, the satellite-based measure, constrained by low spatial resolution and high cloud cover during the flood period, identifies a substantially different set of flooded communities and suggests that flood exposure increases non-farm enterprise and total household earnings. These results imply that harmful flood exposure does not induce broad livelihood diversification in this setting, but that commonly-used flood exposure measures can lead to sharply different conclusions about climate adaptation responses.

Drought and Agricultural Adaptation: Land Reallocation and Labor Dynamics in Vietnam

Giang Thai
,
University of Minnesota
Lifeng Ren
,
AAEA
Khoa Vu
,
AAEA
Nguyễn Dinh Tuần Vương
,
AAEA

Abstract

Climate change increases the frequency and severity of droughts, posing major risks to agricultural production in developing economies. We study how droughts reshape land allocation and input choices in Vietnam, a major rice exporter highly vulnerable to climate risks. Using satellite-based land use data combined with agricultural and firm-level census data, we estimate both short-run and long-run effects through panel fixed effects and long-difference specifications to distinguish immediate adjustments to annual drought shocks from persistent structural change in response to sustained climate change. We document four main findings. First, droughts reduce rice productivity and cultivation area in the long run despite short-run intensification. In the short run, farmers concentrate rice production on smaller plots, raising yield modestly while shrinking harvest area. In the long run, however, droughts exert large adverse effects on rice yield and harvest area. Second, droughts reallocate land away from rice toward other crops and aquaculture. Short run effects are modest but they intensify in the long run. As a mechanism, droughts interact with local labor markets to shape adaptation pathways. We find that droughts affect non-agricultural sectors relatively more than agriculture, pulling workers back into farming. The additional agricultural labor is absorbed primarily in crop cultivation, supporting rice intensification and diversification. Strikingly, aquaculture expansion is driven entirely by machinery adoption rather than labor inflows. Lastly, adaptation patterns are heterogeneous across regions and ecological conditions, reflecting differences in water availability, salinity, and baseline production systems. The results shed light on various adjustment margins to droughts in interaction with local labor market conditions in determining adaptation strategies.

The Welfare Impacts of Large-Scale Land Transactions: Evidence from Ethiopia

Jedidiah Silver
,
Hebrew University of Jerusalem
Yeshwas Admasu
,
AAEA
Dagim Belay
,
AAEA
Workineh A. Kassie
,
AAEA
Melkamu W. Tadesse
,
AAEA

Abstract

Large-scale land transactions (LSLTs) have expanded rapidly across developing countries over the past decades, especially in Ethiopia. While LSLTs are often promoted as instruments for rural transformation, promising employment creation, technology transfer, and improved market integration, they also raise concerns about displacement, reduced access to natural resources, and weakened customary tenure. We combine two analyses to study the medium-run effects of LSLTs in Ethiopia. First, we leverage the plausibly exogenous timing of LSLT entry to estimate the effects on local economic and environmental outcomes. Second, we conduct an original survey of households in Gambela region—an LSLT hotbed—including detailed questions on households’ interactions with LSLTs to shed light on mechanisms. Here we rely on the fact that potential investors relied on observable factors such as soil characteristics, road access, and proximity to conflict zones to identify causal effects using double-debiased machine learning. We find modest effects on crop productivity at the national level with no effect on deforestation, and significant reductions in burned area. In Gambela, we find similar effects on crop productivity and burning but also that LSLTs reduced deforestation and increased tree cover. We estimate a 1.6 percentage point increase in the probability of engaging in wage employment for every 1,000 ha of LSLT within 5–10 km, but no other effects on other labor market outcomes, agricultural production or consumption. These results are consistent with limited labor market and other direct linkages between LSLTs and households and suggest that both the main promises and pitfalls of LSLTs may be overstated in such contexts.
JEL Classifications
  • O1 - Economic Development
  • Q1 - Agriculture