Assessing Values-based Sourcing Strategies in Regional Food Supply Networks : An Agent-based Approach
those values . For an in-depth analysis of the interview findings and the personas developed from them see Krejci et al . ( 2016 ).
Modeling Methodology
In this section , the agents that inhabit the empirically based ABM of the central Iowa regional food system ( producers , consumers , food hub manager ) are described , and an overview of the model and its constituent submodels is provided .
Agents
The consumer agents in the model are described by three parameters : their persona , their demand category , and their food familiarity level . Based on the results of the interviews with food hub consumer members , the probability of the generation of a consumer agent having a given persona in the model was 54 %, 23 %, 5 %, and 18 % for being Locavores , Pragmatists , Frugalists , and Idealists , respectively . Each consumer agent is assigned a demand category , which describes its level of demand ( low , medium , or high ) for each of six product categories in each distribution cycle . The probability that the model generates a consumer agent in any given demand category was determined via food hub historical data , which indicated that many ( 47 %) of its participating consumer members were relatively low-volume customers . Each consumer is also assigned to categories that represent its likelihood of being familiar with a food or finding a food to be “ novel ” in a given interaction with a producer : 50 % of consumers will find 5 % of food interactions to yield foods that are unfamiliar / particularly novel to them , and the other 50 % will encounter unfamiliar / novel foods in 10 % of their interactions . For consumers who prefer familiar foods , the encounters with unfamiliar food will reduce their overall appraisal of the producer who provides it . In contrast , for consumers who prefer novel foods , this type of encounter will increase its rating of a producer .
Consumer agents are also characterized by a utility value , which is a measure of the consumer ’ s satisfaction at any point in time . The higher a consumer ’ s overall utility value is , the more likely he / she is to engage in commerce with the food hub in a given cycle , which influences his membership status . Utility values are scaled from 0 to 1 , with 0 being the least preferred value and 1 the most preferred . The direction of preference for these utility distributions tends to be intuitive ; for example , consumers prefer low prices and highly nutritious / fresh / safe food .
Producer agents are characterized by 11 parameters , each of which governs how the agent is evaluated by consumers and / or how it makes its decisions . Table
31