Closely-associated with calls for evidence-based policy-making, computer models are increasingly used to support public policies on a wide range of complex and controversial issues, ranging from energy and trade policies to regulating chemicals. In this article, we discuss three trends in modelling for policy and their implications for the authority of models in policy-making. First, we observe the proliferation of models and the differentiation of respective modelling landscapes and communities, resulting in intensified competition for political authority and early path dependencies in their political use. Second, different models are increasingly coupled or integrated into complex model frameworks, raising questions of comprehensibility and accountability. Third, modelling for policy experiences calls for transparency and opening that should serve scientific demands of replicability and verifications as well as political demands of legitimacy. In conclusion, we emphasise the continuous relevance of social scientific analyses of modelling for policy.

In this article, we discuss the (re)production of socio-political discourses in two modeling communities, energy transition and transnational trade. Methodologically, we build on bibliometric and qualitative analyses of academic articles. Our analyses show how models structure epistemic communities and are closely linked to specific discourses. The modeling of the energy transition is driven by and contributes to discourses on mitigating climate change and access to energy. Different trade models address either multilateral or regional trade, yet in each case favoring international trade. Overall, we illustrate how the ‘politics of models’ does not only concern their use at the science-policy interface, but is already inscribed in their development, application, and scientific exploitation. These analyses may help experts, policy makers, and the public to better assess the knowledge claims and evidence politics of computer modeling.

In nano risk governance, we observe a trend toward coupling and integrating a variety of computational models into integrated risk governance tools. This article discusses the development and design of such integrated tools as ‘nano risk governance imaginaries in the making.’ Using an illustrative example, the SUNDS tool, we show how the tool manifests conceptual shifts from risk to innovation governance, a technocratic evidence culture based on the quantification of risks, and an envisioned application in industrial innovation management. This conceptualization runs the risk of narrowing the view of nano risks and cementing the widely lamented democratic deficit in risk governance. We therefore conclude that the development and application of integrated governance tools are highly relevant for technology assessment (TA) and TA should actively engage in their development processes.