Background: Global estimates indicate that 25 million people are in situations of ‘modern slavery’ in the form of forced labour or human trafficking, many of whom are migrant workers in low-wage sectors. Recent dialogue has linked ‘unfair recruitment’ with exploitative work outcomes. However, the definitional boundaries of recruiters and recruitment are still opaque and highly context-dependent. Many recruitment initiatives narrowly address legal recruitment and fail to address the circumstances of migrants navigating irregular migration corridors, such as between Myanmar and Thailand, which is the geographical focus of this thesis. This thesis aims to explore how Myanmar migrants make decisions, plan, and navigate migration and the role of mixed social and intermediary networks. Additionally, this thesis describes a novel application of complex systems methods to explore and describe the complex causality of precarity or ‘hyper-precarity’ outcomes among migrants. Methods: This thesis relied on three methodological components, including: 1) a systematic review; 2) mixed methods social network analysis of primary data collected in Thailand; and 3) development of an agent-based model using the primary data from Myanmar migrants. The systematic review synthesised study findings on the use of agent-based modelling in migration and modern slavery research. The mixed methods social network analysis study identified the range of actors, networks, and corresponding pathways in the Myanmar-Thailand corridor. The fieldwork included conducting interviews with 100 Myanmar migrants across three sites in Thailand (Mae Sot, Phang Nga, and Mahachai) using a custom-designed data collection tool to capture egocentric network maps and migration narratives. Finally, the structured social network data and unstructured data from semi-structured interviews with migrants were used to build an agent-based model and produce simulations to observe the emergence in migrant network composition, migration pathways, and migrant precarity outcomes. Results: A broad range of social and intermediary actors influence and facilitate labour migration pathways between Myanmar and Thailand. Myanmar labour migrants experience hyper-precarities in the various pathways arranged by intermediary actors, but migration that is coordinated by family members, which generally involves irregular pathways, was the least precarious in terms of livelihood, immigration, work, and social factors. Findings also show that migrant precarity is not static throughout the journey or even once migrants are at their destination, because there remain numerous decisions and events that can increase or reduce precarity, such as changes in workplace or documentation. The methodological findings indicate that complex systems’ approaches, and specifically agent-based modelling, have not been used to explore low-wage labour migrants’ decision-making or migration pathways – particularly the role of intermediaries. Further, migration ABMs more generally have been limited in their empirical-embeddedness and have rarely drawn on participatory or qualitative methods to inform the model design. The complex system, mixed methods study design implemented in this thesis shows the feasibility and appropriateness of mixed methods social network analysis to inform and validate migration ABMs. Conclusion: This thesis fills a gap in our understanding of mediated migration in the Myanmar-Thailand corridor and how mediated pathways shape migrants’ varied experiences and about the potential roles of intermediaries in precarity-related outcomes. The empirical findings suggest that there are a variety of intermediaries who play diverse roles, which do not conform to many of the stereotypes assumed in safe migration dialogue or interventions. Findings indicate that current assumptions do not reflect the reality of the roles of intermediaries or associated risks for migration outcomes. The empirical contributions of this work challenge current ‘safe migration’ intervention assumptions, especially the simplistic dichotomy of regular/legal versus irregular/illegal migration. Further, the methodological advancements offer a promising way to identify and test intervention assumptions and simulate potential intervention outcome pathways. Ultimately, this work argues that to make migration safer, not just more orderly, we first need to abandon erroneous dichotomies about risk. Once these assumptions are discarded, then we can generate and apply evidence about the complex causal mechanisms that determine labour migration outcomes using data that reflects migrants’ choices, experiences and contextual realities.