Between control and independence: computational modelling within EC’s trade sustainability impact assessments

ABSTRACT Sustainability Impact Assessments (SIA) are a central instrument for evidence-based policy-making in EU trade policy. Computational modelling is the main analytical tool to assess the potential economic impacts of trade agreements. While modelling has long been undertaken by external consultancies, for recent SIAs DG Trade conducts the trade modelling itself. Against the background of this shift from external to in-house modelling, the article addresses the (perceived) roles and authority of modelling in SIAs. Based on the notion of models as boundary objects and two recent SIAs, i.e. TTIP and EU-Australia, we sketch the socio-technical arrangements of models in SIAs. We then discuss the different understandings of the role of modelling by policy-makers, experts, and stakeholders. The in-house shift exposes disagreements on the character and function of models. We further reflect on the potential implications of the in-house shift for the authority of models in SIAs. Our study suggests that there are advantages to in-house modelling, such as control, flexibility, and consistency. However, these might come at the expense of the perceived independence of policy appraisals.