Three Essays on Food Assistance, Environmental Stressor, and Food Choices
Restricted (Penn State Only)
- Author:
- Feng, Jinglin
- Graduate Program:
- Energy, Environmental, and Food Economics (PhD)
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 22, 2023
- Committee Members:
- Daniel Brent, Outside Field Member
Paul Grieco, Outside Unit Member
Douglas Wrenn, Program Head/Chair
Edward Jaenicke, Major Field Member
Linlin Fan, Chair & Dissertation Advisor - Keywords:
- dietary quality
SNAP
instrumental variables unconditional quantile regression
caloric acquisitions
charitable food assistance
SNAP benefit month
Healthy Eating Index
air pollution
wildfire smoke exposure
food purchases - Abstract:
- This dissertation consists of three essays on food assistance, environmental stressor, and food choices: (i) The first essay analyzes the distributional impact of the Supplemental Nutrition Assistance Program (SNAP) on diet quality. (ii) The second essay examines heterogeneity in the use of social networks and charitable food assistance over the SNAP benefit month. (iii) Lastly, the third essay focuses on the effect of short-term ambient air pollution on diet quality, using household scanner data. SNAP is the nation’s largest domestic food and nutrition assistance program for low-income Americans. Recent studies that examined the effect of SNAP on diet quality focus on the average effects. In essay one, we use the 2012 USDA’s National Household Food Acquisition and Purchase Survey (FoodAPS) data and an unconditional quantile estimator to examine the distributional impacts of SNAP on diet quality, as measured by Healthy Eating Index-2010 (HEI-2010). To identify the differential impacts of SNAP across the distribution of diet quality, we exploit exogenous variation in state’s maximum weekly unemployment insurance (UI) benefits and state outreach spending per capita as instrumental variables (IVs). We find that SNAP has no significant impact on households’ diet quality on average. However, for households with initially low-to-intermediate diet quality, SNAP participation reduces their HEI scores by over 17% or more than 7 points out of a total score of 100. The negative impacts of SNAP on these HEI quantiles are mainly driven by an increased acquisition of empty calories. As low-income households often combine personal resources with both public and private food assistance in times of need, understanding how they fulfill their energy needs over the SNAP benefit month is crucial. In essay two, we use the 2012 USDA’s FoodAPS data to examine the strategies SNAP households use to meet their energy needs throughout the benefit month, focusing on complementary food sources like social networks and charitable food assistance (CFA). We also explore heterogeneity in the use of these food sources. Our study yields three key findings. First, we find a significant spike in calorie acquisition on benefit receipt day (day 0), rather than a SNAP cycle throughout the month. Second, both social networks and CFA play an important role in food acquisition, particularly for households not owning a vehicle, albeit homeownership and income moderate this impact. Third, diet quality does not change over the course of the benefit month. Air pollution, as the largest environmental health risk factor worldwide, has well-documented adverse effects on human health and well-being, yet its impact on consumer food choices and diet quality remains largely unknown. Essay three studies the causal effects of short-term ambient air pollution on diet quality, as measured by Healthy Eating Index-2015 (HEI-2015), using wildfire smoke exposure from 2010-2018 as a source of exogenous variation for air pollution. By linking nationwide satellite-based smoke plume data, ground-based pollutant measurements, and consumer scanner data for more than 120,000 U.S. households, we find no impacts of air pollution on overall diet quality or individual diet components. This suggests that air pollution levels might not be a substantial driver of household dietary choices. Our findings reveal a socioeconomic gradient in diet quality, with lower-income households, less-educated household heads, and counties with higher PM2.5 levels consistently exhibiting poorer diet quality. Moreover, we observe no evidence that the effects of air pollution vary across income, education, and county pollution levels.