Association for Public Policy Analysis & Management

Panel: Food Choices and Vulnerable Populations

(Poverty and Income Policy)

In this panel we explore how environmental factors and policy can affect food choices that vulnerable populations make. The four papers in the panel explore how children from low-income households, people living in food deserts, recently released convicts and households during the Great Recession make decisions about what to consume, how much to consume and the means to consume. The papers highlight ways that policy can lead to both positive and negative food-related decisions as well as bring to light new vulnerabilities that households have that impact these decisions.

The first paper presents the results from a field experiment that tests the effectiveness of information prompts in encouraging children in a low socio-economic area school to make healthy food choices in their National School Lunch Program meals. The authors find that simple verbal prompts increase the proportion of children choosing and consuming white milk relative to sugar-sweetened chocolate milk by 10 percentage points. These results inform policy makers of a simple and cheap tool in promoting healthy consumption among children who are at a high risk for obesity. The second paper analyzes the effectiveness of a new program that delivers food to people living in low food access areas of Baltimore City. Using surveys, the author finds that program participants reduced their usage of supermarkets and restaurants as well as increased their purchases of healthy foods after the introduction of the delivery program. The author further investigates these claims by comparing the composition of program purchases from administrative data to those found in the USDA's National Household Food Acquisition and Purchase Survey. This paper provides evidence of a program that is able to increase healthy food purchasing by increasing access to healthy foods. The third paper uses administrative dataset from a corrections agency to analyze the effect of a policy that places a lifetime ban on some convicts from ever receiving SNAP. Using a regression discontinuity design the author finds that the SNAP ban increases recidivism among released drug traffickers where the increase is primarily driven by crimes that are financially motivated. These findings highlight how access to nutritional assistance programs can dramatically impact the trajectories of some populations. The final paper investigates why food insecurity increased during the Great Recession even when one controls for the income-to-poverty ratio of households. The authors use Current Population Survey data to find that food insecurity rose the most in places that were hit hardest by unemployment duration and drops in housing prices. Hence the authors conclude that, during the Great Recession, people lost access to resources that previously allowed them to consume more food and that these resources are not captured by the income-to-poverty ratio measure.

All of these papers employ multiple data sources in supporting their findings including some combination of self-collected data, survey data, administrative data and national surveys. The utilization and introduction of new datasets in this panel complements the themes of this conference of "Better Data for Better Decisions."