��������������������������
may not actually be at fault for contaminating a food product , in this model if they serve the product to customers they bear responsibility . ������������������ : As noted , people who are elderly , very young , pregnant , or immunocompromised are more prone to foodborne illnesses and often suffer more severe consequences should they become ill ( Gerba , Rose , & Haas , 1996 ). In this model , 30 % of consumer agents were designated as being at risk . This level was based on 2011 Canadian census data on the number of Canadians five and under , and over age 60 ( Statistics Canada , 2013 ). The number of pregnant women in the model was extrapolated from the census using the number of children under the age of one . The number of individuals suffering from diseases that contribute to immune-compromised status were available prior to 2011 but not in a useful format . To avoid double-counting , this data was excluded , but the final level used in the model was rounded up to 30 % from the prior estimates based on age and pregnancy ( see Table 1 ). 6 We recognize that this estimate is incomplete and that the total number of at-risk individuals in Canada is likely to be even higher .
Table 1
At Risk Population Total Canadian Population 33,476,690 Adults over 60 6,997,725 Children 5 and under 2,241,840 Pregnant women 367,930 Total at risk 9,607,495 Percentage of population 28.70 %
Table 1 . ������������������
In addition , exactly what level of contamination in a food product is unsafe and will thus translate into illness is unclear ( Rawson & Becker , 2004 ). Therefore , any parameter for the probability of contracting an illness if a consumer goes to a contaminated restaurant is necessarily difficult to justify . All else equal , the at-risk population is more likely to become ill from exposure to a contaminated restaurant . In this context , the probability of an at-risk consumer becoming sick from exposure is set at 0.3 whereas it is assumed that the rest of the population has a probability of 0.15 of becoming ill given exposure . Additionally , at risk consumers remain sick for five time steps as opposed to three , to reflect the more severe consequences of foodborne illness for high risk individuals . ��������� ��������� : Restaurant food is an experience good in that its quality cannot be ascertained prior to consumption ( Henson et al ., 2006 ). One survey that investigated consumers ’ perception of restaurant food safety found that 56.4 % of consumers stopped frequenting a restaurant based on food safety concerns ( Henson et al ., 2006 , p . 285 ). In addition , research conducted by the Food Standards Association in the United Kingdom indicates that , if individuals had concerns about hygiene , up to 70 % of respondents would not purchase again from a food service outlet ( as cited by Choi , Nelson , & Almanza , 2011 ). In this model , this concept of avoidance has been incorporated in a simplified way . Consumers that have become sick simply do not return to that restaurant . While a bit extreme , it does capture the notion of consumer memory in a simplified way . This assumption is
6
Author ’ s calculations based on The Canadian Population : Age and Sex from Census 2011 ( Statistics Canada , 2013 ).
145