Over the past 28 months, ON-MERRIT has collected evidence on how Open Science reshapes scientific endeavours and their interaction with societies and industries, conducting interviews, surveys, group discussions, document analysis and the analysis of large bibliographic databases. While a host of empirical results are now available via several comprehensive deliverables, the ON-MERRIT consortium is conducting a final and key step: synthesizing the knowledge gained and distilling it into actionable recommendations. The final recommendations will be based on the evidence collected so far, and refined in light of highly productive discussions with domain experts. Still, assessing the recommendation���s efficacy and any unintended consequences based on the relationship ���evidence about system A��� ��� policy is hard. To close the circle by investigating the relationship policy ��� ���change in system A���, this deliverable leverages the method of Agent-Based Modelling (ABM) to assess outcomes of potential policies by modelling patterns of the uptake of Open Science (OS) under policies proposed in the research literature. Data sharing has become an integral part of the OS agenda, and is central to an increasing number of policies. The sharing and subsequent re-use of research data is expected to lead to substantial efficiency gains across science and industry (Directorate-General for Research and Innovation, 2016). There is substantial evidence that sharing of research data leads to a citation advantage (Colavizza et al., 2020; McKiernan et al., 2016). Game-theoretical approaches (Pronk et al., 2015) have highlighted that, as with many other aspects of OS, the overall adoption and uptake of data sharing can be understood as a problem of collective action (see also Scheliga & Friesike, 2014). That is, system-level benefits can be expected when sharing is the norm, while individuals may benefit from not sharing their own data while re-using data that others have shared (i.e., free-riding). In this deliverable, we develop and discuss the DASH (DAta Sharing) model, which leverages the potential of ABM to investigate how plausible scenarios of incentivising data sharing could encourage its uptake, given varying scenarios of costs associated with the sharing of data, as well as different strategies employed by individual research groups to decide on whether or not to share data. The model serves as a starting ground, enabling further evaluation of policy interventions for their efficacy, and potential repercussions for equity in science.