Studies that assess the effectiveness of the Supplemental Nutrition Assistance Program (SNAP) in improving food security have primarily been conducted using Federal survey data. These surveys have high rates of SNAP misreporting. Research has shown that even small amounts of misreporting could impact estimates of the causal and noncausal relationships between SNAP participation and food security. In this context, we use administrative records from 19 States matched to the Current Population Survey (CPS) Food Security Supplement for the 2017–19 period to investigate misreporting and its consequences for understanding how SNAP impacts food security. SNAP underreporting is about 47 percent in these years, comparable to studies using previous years of CPS microdata. At the same time, overreporting is slightly higher than some other studies at 2 percent. We find the administratively matched SNAP sample is more food secure than the sample based on survey responses only. Our results indicate that SNAP has a potentially large and positive effect on food security. However, the estimates of the effect of SNAP on food security for the administrative sample are more precise than for the survey sample; estimates from the survey sample likely overestimate the efficacy of SNAP.
Studies that assess the effectiveness of the Supplemental Nutrition Assistance Program (SNAP) in improving food security have primarily been conducted using Federal survey data. These surveys have high rates of SNAP misreporting. Research has shown that even small amounts of misreporting could impact estimates of the causal and noncausal relationships between SNAP participation and food security. In this context, we use administrative records from 19 States matched to the Current Population Survey (CPS) Food Security Supplement for the 2017–19 period to investigate misreporting and its consequences for understanding how SNAP impacts food security. SNAP underreporting is about 47 percent in these years, comparable to studies using previous years of CPS microdata. At the same time, overreporting is slightly higher than some other studies at 2 percent. We find the administratively matched SNAP sample is more food secure than the sample based on survey responses only. Our results indicate that SNAP has a potentially large and positive effect on food security. However, the estimates of the effect of SNAP on food security for the administrative sample are more precise than for the survey sample; estimates from the survey sample likely overestimate the efficacy of SNAP.