This article examines how households adjust food purchasing behavior following a type
2 diabetes diagnosis, using the Nielsen Consumer Panel (2010-2023) merged with the
PanelView survey (2011-2023). Combining Callaway and Sant’Anna (2021) staggered
di!erence-in-di!erences estimator with a reference price framework that decomposes
prices into loss and gain domains (Caputo et al., 2020), I analyze category-level purchasing
responses across fresh produce, non-fresh produce, milk, soft drinks, and sugarsweetened
beverages (SSBs). The diagnosis triggers a significant and persistent 5-6
percent reduction in non-fresh produce purchases across the full sample. Income heterogeneity
analysis reveals a striking gradient, as high-income households (above $40,000)
reduce SSB purchases by 10-15 percent and restructure multiple food categories simultaneously,
while low-income households concentrate adjustments on non-fresh produce
alone. Expenditure-based classification provides complementary evidence of limited
adjustment flexibility among the lowest-spending households. These findings indicate
that the effectiveness of health information as a behavioral nudge depends critically on
economic capacity, with implications for the design of dietary interventions targeting
diabetic populations.
