How Do People with Autism Spectrum Disorder (ASD) Perform on the Ultimatum Game? Using Game Theory to Study Social Decision Making in ASD

Background:   This research seeks to elucidate how people with ASD behave in real-world social situations, by studying social decision making, herein defined as the capacity to attend to, process, interpret and act upon information to inform spontaneous, effective interpersonal behaviour. Game theory researchers working in behavioural economics have developed psychometrically sound, ecologically valid tools for measuring social decision making, but these have rarely been used to study the strengths and difficulties of people with ASD. This study investigated social decision-making in ASD using a game theory task called the Ultimatum Game (UG), which is an interactive bargaining task measuring reciprocal responses to social norm violations and unfairness. In the UG, a sum of money, for example £10, is to be divided between two players – William and Mary. William chooses how to share the money, for example making a fair offer (£5) or an unfair offer (e.g. £2) to Mary. Mary can then decide whether to accept or reject William’s offer. If she accepts, both parties will each receive the agreed share of the money. If she rejects, neither will receive anything. The “ultimatum” refers to the take-it-or-leave-it offer in the game. 

Objectives:  We sought to discover whether young people with high-functioning ASD showed atypical performance on the UG, compared to controls, both in terms of the proposals they made, and their decisions about whether to accept unfair offers. We also investigated whether elements of cognition associated with ASD (namely, difficulties with theory of mind, emotion regulation and executive function) predicted suboptimal performance on the UG. 

Methods:   Twenty young people (aged 11 to 17 years) with ASD and an IQ in the normal range participated in the UG, which was delivered as a computerized task. Their behaviours as proposers of offers, and their responses to offers (including unfair offers) were measured. Participants also completed measures of theory of mind, emotion regulation and executive function, and parents completed a questionnaire on EF.  Two typically developing comparison groups were used, one matched on chronological age (n=20) and the other matched on developmental age (n=20).

Results:  Young people with ASD proposed significantly fewer fair offers to their opponents in the UG, but did not differ from controls in their responses to unfair offers. In the ASD group, participants with better theory of mind were more likely to propose fair offers. Participants with better executive function – especially those with better emotional control and behavioural regulation – accepted more unfair offers. 

Conclusions:  Young people with ASD demonstrated preserved aspects of SDM in a 'reactive' social context, which appeared to recruit executive functioning, especially emotional control and behavioural regulation. They differed to their TD peers when behaving in a 'proactive' social situation, which seemed to depend partly on theory of mind abilities.