Nowcasting Food Insecurity
There are different types of food insecurity. During the COVID-19 pandemic, we saw an increase in the emergent need for food assistance. Emergent need is different from recurrent need in that it is driven by unexpected events, may involve previously unaffected populations, and can be acute and temporary. In spring 2021, the CFIE agreed to fund a research project proposed by Drew Starbird and Ram Bala to predict emergent food insecurity using online
search behavior. The objectives of the project, “Nowcasting Food Insecurity in Silicon Valley,” are (1) to characterize the online behavior of individuals in need of food assistance, (2) connect that information to the quantity and quality of food need in vulnerable communities, and (3) create a dashboard that represents the current and immediate need for food assistance at the zip code level in Silicon Valley.
Preliminary results show a significant correlation between online search behavior and the emergent need for food assistance, as the graph below illustrates. It shows the sum of the indices for four search terms (“calfresh”, “food bank”, “food pantry”, and “supplemental nutrition assistance program”) compared to the total number of SNAP applications for the San Francisco Bay Area between January 2019 and December 2021.
The next step in this project is to collect additional data regarding search behavior and SNAP applications for other geographic regions. This information will allow us to validate the correlation between the search terms and the need for food. Subsequently, we will use the measured relationship to develop the dashboard API.