Lebanon, hosting a large refugee population from Syria and Palestine, faces significant challenges in providing public services. Poor mental health is prevalent among both refugees and the local population, potentially impacting decision-making and healthcare utilization. The study explores the connections between mental health (measured by subjective well-being), risk-taking behavior in financial choices, and trust in institutions, comparing three distinct populations: Lebanese nationals, Syrian refugees, and Palestinian refugees. The researchers hypothesize that poor mental health will be associated with suboptimal decision-making, and that behavioral interventions such as nudges and boosts can improve decision-making. The context of Lebanon's humanitarian crisis highlights the importance of understanding these interrelationships to inform effective policy interventions.

The introduction cites existing research linking poor mental health to suboptimal decision-making. It also notes the varied impacts of displacement on risk aversion, citing studies showing both increased risk aversion due to trauma and increased risk-seeking behavior due to material losses. The World Health Organization's Mental Health Action Plan is referenced as advocating for context-specific approaches. Behavioral insights, specifically nudges and boosts, are presented as promising policy intervention methods, citing previous successful applications in similar contexts like Serbia. The study builds upon this literature by testing a well-being measure in the Lebanese context and investigating the effectiveness of nudges and boosts in improving advantageous choices among diverse populations.

Data were collected online in Arabic from 754 participants (Lebanese, Syrian, and Palestinian refugees) between October and November 2019. A 10-item well-being measure was used, along with decision-making tasks involving choices between risky and safe financial options (both gain and loss framed). Participants were randomly assigned to one of three groups: a control group, a disclosure nudge group (providing expected values), and a boost group (providing detailed instructions on calculating expected values). Confirmatory factor analysis was used to validate the well-being measure. Linear and logistic regressions were used to analyze the relationships between well-being, risk-taking, trust, and the effectiveness of interventions. Power calculations were performed a priori to determine the appropriate sample size. Ethical approval was obtained from the Centre for Business Research, Judge Business School, University of Cambridge and the Lebanese Ministry of Public Health.

The 10-item well-being measure showed similar psychometric properties across all three groups (Lebanese, Syrian, and Palestinian refugees), validating its use in this context. No significant differences in average well-being were found between the Lebanese and refugee populations. However, higher subjective well-being was significantly associated with more risk-taking among refugees, particularly in gain-framed scenarios. This relationship was weaker or non-existent for the Lebanese nationals. Regarding trust, Palestinians reported the lowest overall trust in others, while trust in family and neighbors was higher across all groups. There was no significant relationship between well-being and overall trust. Behavioral interventions had a moderate effect on improving advantageous choices, but this effect was predominantly observed among Lebanese participants in the boost condition. The disclosure nudge did not consistently improve advantageous choices. For the loss frame, the boost condition was particularly effective for those who made advantageous choices at baseline. In contrast, the disclosure nudge led to significantly fewer advantageous choices for Palestinian refugees compared to Lebanese participants.

The study's findings highlight a crucial distinction in the relationship between well-being and risk-taking between refugees and the Lebanese host population. The positive association between well-being and risk-taking among refugees may reflect a greater willingness to seek opportunities to improve their circumstances, even if risky. The lack of similar association among Lebanese nationals suggests that cultural and contextual factors influence this relationship. The general similarity in mental health between local and refugee populations does not necessarily negate the need for parallel healthcare systems, as low trust among refugees may still hinder their uptake of services. The effectiveness of boosts, particularly among Lebanese participants, suggests that interventions supporting autonomous decision-making may be effective in middle-income countries. The lack of effectiveness of nudges and boosts for refugee groups could relate to different data collection methods or the need for more robust decision support tailored to their specific challenges.

This study provides valuable insights into the complex interplay between mental health, risk-taking, and trust in the context of a humanitarian crisis. The findings underscore the need for tailored policy interventions considering the specific needs and vulnerabilities of refugee populations. Future research should explore the generalizability of these findings across a wider range of risk domains, investigate the reasons for the differential effectiveness of interventions, and utilize more ecologically valid methodologies.

The study's limitations include the use of hypothetical financial choices, which may not fully reflect real-world risk preferences, particularly in contexts of extreme stress and uncertainty. The different testing environments (online vs. in-person) could have introduced some bias, although this is likely limited. The lack of data on other potential socio-economic, legal, structural, or cultural factors influencing risk-propensity prevents thorough investigation of complex relationships. The sample size limitations affected the depth of exploration of these relationships. The limited effectiveness of interventions for the refugee groups may be linked to variations in presentation modalities or the complexity of the boost/nudge.