Berlin-based economic research institute applying econometrics and machine learning to energy policy, health economics, and social policy across Europe.
DIW Berlin is one of Germany's leading economic research institutes, applying rigorous quantitative methods — econometrics, machine learning, structural modeling — to real-world policy questions. Their H2020 work spans two distinct domains: energy system economics (demand response, grid integration, public procurement for green technologies) and socio-economic research (health economics, gender pay gaps, early childhood education). They bridge the gap between academic economics and actionable policy advice, providing data-driven evidence that informs European energy transitions, public health strategies, and social policy design.
RealValue (smart thermal storage and grid integration), SET-Nav (clean energy roadmaps), Open ENTRANCE (energy transition modeling), and XPRESS (green public procurement) form a consistent energy economics portfolio.
PEARLE studied early childhood education group composition effects while PAGE analyzed paternity leave impacts on gender pay gaps — both led as coordinator using individual-level micro data.
ABRSEIST (coordinator, EUR 904K ERC Starting Grant) applies Bayesian learning, machine learning, and field experimentation to antibiotic resistance and pharmaceutical demand — a significant new direction.
InGRID-2 contributes to European research infrastructure for inclusive growth, reflecting DIW's role as a major provider of panel survey data (notably the German SOEP).
ABRSEIST explicitly lists machine learning, Bayesian learning, and structural modeling; these methods also underpin their energy modeling work in SET-Nav and Open ENTRANCE.
DIW's early H2020 work (2015–2017) concentrated on energy system economics — smart thermal storage, demand response, grid integration, and energy transition roadmaps. From 2018 onward, the focus shifted markedly toward applied microeconomics and health economics, with ERC-funded research on antibiotic resistance using advanced computational methods (machine learning, Bayesian learning) and continued work on social policy topics. The energy strand did not disappear but became more policy-oriented (green procurement, open energy models), while the analytical toolkit grew more sophisticated.
DIW is moving from descriptive energy economics toward computationally intensive applied microeconomics — particularly health and pharmaceutical markets — making them a strong partner for projects combining economic modeling with machine learning.
DIW operates comfortably in both leadership and partner roles, coordinating 3 of 8 projects (notably two MSCA fellowships and one ERC grant — all individual-researcher-driven formats). As a participant, they join mid-to-large consortia (72 unique partners across 8 projects), contributing economic analysis rather than technical implementation. Their wide partner network (22 countries) suggests they are easy to integrate into new consortia and valued for adding the economic dimension to technically-focused projects.
DIW has collaborated with 72 unique partners across 22 countries, indicating a broad European network rather than concentration in a few bilateral relationships. Their partnerships span energy research institutes, universities, and policy organizations across the EU.
DIW stands out because it combines deep economic research capacity with practical policy orientation — they don't just model energy markets, they quantify the economic impact of policy interventions. Unlike technical energy labs, DIW brings the socio-economic lens that EU projects increasingly require for impact assessment and policy recommendations. Their recent pivot into health economics with advanced ML methods positions them at an unusual intersection of economics, data science, and public health that few European research centers occupy.
