Measuring and testing vulnerability to food insecurity for prediction and targeting

Area of Science:

• Agricultural Economics
• Development Economics
• Food Security Studies

Background:

• The COVID-19 pandemic exposed vulnerabilities in household food security.
• Assessing household vulnerability to food insecurity is crucial for effective policy interventions.
• Existing theory-based methods for targeting food insecurity require empirical validation.

Purpose of the Study:

• To empirically test the validity of theory-based methods for targeting household vulnerability to food insecurity.
• To compare the performance of a theory-based measure against a data-driven routine.
• To analyze the impact of the COVID-19 shock on food insecurity prediction models.

Main Methods:

• Utilized longitudinal survey data from Nigeria (2010-2020) from the World Bank.
• Applied an empirical test to assess a theory-based measure of household vulnerability to food insecurity.
• Conducted sensitivity tests using pre-pandemic data to isolate the shock's impact.

Main Results:

• The theory-based vulnerability measure significantly underperformed in predicting food-insecure households out-of-sample.
• A simple, data-driven routine demonstrated superior predictive performance.
• Poor forecasting was not solely attributable to pandemic-induced data discontinuities.

Conclusions:

• There is a critical need to enhance the effectiveness of targeting approaches for identifying food insecurity hotspots.
• Predictive vulnerability models are constrained by data availability, especially in data-scarce environments.
• Greater methodological innovation is required to improve the anticipation of vulnerability to economic and health shocks.