A Multi-agent Model of Deceit and Trust in Intercultural Trade

Abstract

Abstract. Trust is a sine qua non for trade. According to transaction cost economics, a contract always offers some opportunity to defect. In the case of asymmetric product information, where the seller is better informed about product quality than the buyer is, the buyer either has to rely on information provided by the seller or has to check the information by testing the product or tracing the supply chain processes, thus incurring extra transaction cost. An opportunistic seller who assumes the buyer to trust, may deliver a lower quality product than agreed upon. In human decisions to deceive and to show trust or distrust toward business partners, issues like morality, shame, self-esteem, and reputation are involved. These factors depend strongly on trader's cultural background. This paper develops an agent model of deceit and trust and describes a multi-agent simulation where trading agents are differentiated according to Hofstede's dimensions of national culture.

Computational Modeling of Culture’s Consequences

“…It has been validated in experiments with Dutch adolescents and adults [18]. If no validated model can be found in literature, a dedicated model has to be formulated based on empirical data or research; see, for instance, [19].…”

“…In the example of trade the following have been identified as relevant relational attributes: status, in-group versus out-group membership, and the trust relation between partners. For instance, the vector of cultural factors influencing the decisions to deceive and to trust identified by Jonker et al [19] can be taken from the column labeled "Cultural factor" in Table 2. …”

“…By the knowledge acquisition process taken, the increasing and decreasing effects of the cultural factors can be identified, as illustrated in Table 2 [19]: L i + is the set of factor labels that have a + sign in the column associated with the parameter labeled p i ; L i − is the set of factor labels that have a minus sign in the column associated with p i . The next problem to solve is the combination of these influences into a single effect on each parameter, i.e.…”

A Review of Theoretical and Practical Challenges of Trusted Autonomy in Big Data

“…• Detecting Deceitful Agents: Deceit is considered commonplace in certain industries such as trade [46], [47], [48], [49], computer networks [50] and online communities [33]. Understanding deceit is vital for detecting it, and for the design of governance mechanisms to discourage deceit.…”

Trustworthy Organic Computing Systems: Challenges and Perspectives

“…This fact is witnessed by the strong emphasis on relationships in the literature [45], but also made explicit in some definitions of trust [17]. Cultural differences can also heavily affect the disposition to trust or deceive [21]. As the final decision about whether or not to trust is an individual one and can be based on very different criteria, different general attitudes and different experiences, trust is also not a transitive property that is inherited or passed on.…”