This study assessed implementation of food defense practices in public schools in Montana, Wyoming, South Dakota, North Dakota, Iowa, Minnesota, and Wisconsin. The first phase involved a qualitative multi-site case study: one-day visits were made to five school districts in the states of Iowa, South Dakota, Minnesota, and Wisconsin. A principal, district foodservice director (FSD), two food production workers, and an emergency responder at each site were interviewed about food defense awareness and risk perception. Meal production and service were observed for implementation of food defense practices. In Phase Two 543 school food authorities or FSDs (36% percent of the population from 1,501 districts in seven Midwestern states) responded to an Internet-administered survey. Survey items included frequency of implementation of 31 food defense best practices adapted from the work of Yoon and Shanklin (2007a) and Yoon (2007). The survey included ten items assessing risk perception using Slovic's psychometric paradigm (1987). Items requested information about crisis management and food defense planning, food defense training, influence over districts' security policies, as well as operational and demographic characteristics.

Four themes emerged from the 25 interviews conducted during the site visit: low awareness, lack of concern, food not considered a potential danger, and how conflicting priorities influence security. Food defense was an unfamiliar concept among most interviewees. Many expressed the belief that food tampering was not likely in their schools because employees were trustworthy or location was too insignificant. Principals expressed concern for physical security measures but did not perceive their contribution to food defense. In most districts, the FSD was not included in district emergency response planning activities and communication about food defense did not occur between principals, FSD, and emergency responders. Some of the interviewees had experience with food tampering incidents; seven incidents were reported, of which five had occurred in schools. Employees in a central kitchen facility were suspected in two of the school incidents, and three were perpetrated by students, indicating different sources of vulnerabilities.

Most (67.8%) survey respondents reported district enrollment <2,500 students. Few (14.5%) had implemented a food defense plan; implementation was related to FSD involvement in crisis management planning and to FSD receiving food defense training. Thirteen practices were implemented most of the time (mean >4.0 on 5-point scale with 5 = always); most of these within control of FSD. Six practices were implemented less frequently (mean <3.0 on 5-point scale with 1 = never); three would require administrative action to implement, and two were related to FSD communication with emergency responders.

Mean values for "unknown risk" risk perception measures indicated some disagreement that intentional food contamination was a new risk for respondents and strong disagreement that they personally knew a lot about how terrorists could contaminate the food supply. The mean for the dread risk scale was 1.93 on a 4-point scale with 4 = high, similar to perceived risk of common everyday activities reported by Lee, LeMyre, and Krewski (2010). Compared to district administrators, FSDs perceived significantly greater personal control over both terrorism and food tampering risks.