Submissions Accepted for Presentation at the World Bank Land Conference 2024

In a world with increasing climate change and intensifying food insecurity, agricultural productivity is central to development. Increasing agricultural productivity requires a detailed understanding of cultivated land and its limitations, and how these shortcomings can be addressed. In this regard, both land quality and quantity play a critical role. The ability to appropriately estimate the degree to which these inputs positively or negatively affect production is dependent on accurate measurement. With an evolving technological landscape, the menu of tools available to potentially measure land quality and quantity, particularly in cost-effective and scalable ways, is expanding. Through a methodological survey experiment implemented in Uganda, we set out to validate innovative approaches for measuring area (land quantity) and soil health (land quality) and assess their feasibility for implementation in household survey contexts as well as the policy-relevant implications of their use.

While monitoring soil quality provides indispensable inputs in agricultural policies in Africa, it is expensive and a logistical challenge to collect high-quality data on soil quality from the traditional household survey. We offer an early assessment of an alternative, less expensive imputation-based method to address this data gap. The central idea is to leverage a smaller benchmark sample with high-quality soil data—in combination with a larger survey without any soil quality data (or with low-quality soil data)—to generate imputation-based estimates in the larger survey. Preliminary results for Ethiopia and Uganda are very encouraging, suggesting that imputation-based estimates are reasonably close to the estimates based on the benchmark surveys. If replicated in other contexts, including for other agricultural variables, these results could open up a new and cost-effective way to address the challenge of missing soil quality data for Sub-Saharan African countries.

We assess the reliability of measured farm sizes (ownership holdings) in the Living Standard Measurement Study – Integrated Surveys on Agriculture (LSMS-ISA) in Ethiopia and Malawi based on three survey rounds (2012, 2014, 2016) in Ethiopia and four rounds (2010, 2013, 2016, 2019) in Malawi. By using the balanced panel of households that participated in all the rounds, we utilized the within-household variation in reported and measured ownership holdings that were mostly measured with GPSs and/or with rope and compass. While this gives reliable measures of reported holdings, we detect substantial under-reporting of parcels over time within households that largely have been overlooked in previous studies. We find that the estimated farm sizes within survey rounds are substantially downward biased due to systematic and stochastic under-reporting of parcels. Such biases are substantial in the data from both countries, in all survey rounds, and in all regions of each country.

We report the results of two survey experiments designed to shed light on a persistent mystery in rural household survey data from Africa: why there are so many fewer self-reported landlords (renters-out) than tenants (renters-in).

We find that nudging induces a significant increase in the reported rate of renting-in parcels but has negligible effects on reported rates of renting-out and sharecropping out. However, our list experiment indicates much higher revealed rates of renting out (14-15%) than is reflected in the nominal plot-roster responses (3%). The magnitude of the latter finding fully explains the apparent difference in renting in vs. renting out rates derived from the regular parcel roster responses. Our results demonstrate that simple nudged embedded in survey designs can improve reporting rates in land market participation while the list experiment findings hint the need for specialized survey designs to understand true level of land market participation in similar contexts.