The frequency of affirmative responses to food security items falls with increased total income and is higher for food stamp program participants.

An official goal of the Food Stamp Program (FSP) is to reduce household food insecurity, which the federal government measures using an 18-item survey module. This study combines features of "dose-response" and "participant/non-participant comparison" research designs to examine the impact of total income on individual food security items among 10,175 FSP participants and 144,372 non-participants using data from the 2001-2004 Current Population Surveys (CPS). We calculated the log odds of affirming 8 adult-specific items: worried food would run out, food bought didn't last, couldn't afford balanced meals, an adult cut or skipped meals, ate less, went hungry, lost weight, or did not eat for a whole day. FSP participants had higher odds of affirming each of the 8 items. As total income increased, the odds of affirming each of the 8 food security items decreased for FSP participants and non-participants. In addition, we measured "participant/non-participant difference" for a particular item as the log odds of affirmative response for participants at a particular total income level minus the log odds of affirmative response for non-participants at the same total income level. Across all income levels, the participant/non-participant difference was greatest for the least severe food security items, "worried food would run out" and "food bought didn't last" and smallest for the most severe, "did not eat for a whole day". [ABSTRACT FROM AUTHOR]