Aims and Objectives
Agent-based models (ABM) are a powerful and increasingly essential tool in economics and across other social sciences for studying complex systems and social challenges. This summer school addresses the critical need for researchers to master not only the conceptual underpinnings of ABM but also the practical challenges of programming, adhering to robust protocols, and effectively communicating model design and results.
Building on past successful editions, this intensive, full-time course equips participants with the advanced theoretical understanding and skills to design, implement, validate, and analyse robust agent-based models, such as a PhD thesis, a high-quality academic paper, or a policy report.
In this edition we will also evaluate the power and limitations of Large Language Models (LLM) as co-pilots in designing and coding ABMs. In an era where researchers increasingly turn to LLM to generate code, this course emphasises the critical importance of understanding model mechanics. By comparing AI-generated models with manual coding, students will appreciate the need for bridging theory and implementation.
While the methodological skills acquired are applicable across all social sciences, the primary case studies will focus on economics, innovation, development, political economy, public policy and structural change.
What you will Learn
Upon completion, participants will be able to:
- Understand the foundational principles and theoretical frontiers of ABMs, including their strengths and limitations in social science research.
- Design and implement ABMs from scratch in Laboratory for Simulation and Development (LSD) and R, abstracting away from tedious software engineering to focus purely on modelling logic.
- Analyse model output rigorously and apply Machine Learning methods for advanced sensitivity analysis with R.
- Develop skills useful to produce ABMs also in other platforms/languages such as Python and NetLogo.
- Critically compare AI-generated ABMs with expert manual coding in R and LSD, recognizing the methodological risks of treating simulation code as a black box.
- Critically evaluate ABM methodology, navigate the pitfalls of relying on AI for model generation, and apply this learning to their own research questions.
- Conceptualise and prototype an original ABM project, laying a strong foundation for future development and publication.
Course Structure
The course combines expert-led theoretical lectures, hands-on coding exercises under the close supervision of dedicated teaching staff, interactive group discussions, and seminars showcasing advanced ABM applications by leading researchers. Participants will work on individual and group research projects throughout the week, culminating in presentations of their model prototype on the final day. Dedicated time each afternoon is reserved for individual consultations with faculty offering personalized feedback on your research ideas and projects.
Day 1: Methodological foundations, theory, and conceptualizing ABMs in economics and the social sciences
Day 2: Developing and implementing ABMs in R (manual vs. AI-generated coding) and analysing models in Python
Day 3: Overcoming coding skills limits with Laboratory for Simulation Development (LSD)
Day 4: Broader social science applications (Policy applications, SDGs, pandemics, green techs); analysing models in NetLogo; and group hackathon.
Day 5: Sensitivity analysis, validation techniques, hackathon completion, and student presentations.
Day 1-5: 1-on-1 consultations with faculty and facilitated group discussions
Who Should Apply?
This course is designed for:
- Ambitious graduate and PhD students in economics and other social sciences with a keen interest in advancing their research using computational methods.
- Early-career researchers and experienced modellers looking to enhance their skills and explore new frontiers in ABMs.
We particularly encourage applications from individuals eager to integrate computational modelling into their social science toolkit. Prior experience in programming can be helpful but is not required. The course welcomes and supports participants with limited or no prior programming experience, providing them with the tools needed to model without having to learn hardwired code.
Faculty: Learn from World-Leading Experts
The school’s faculty includes renowned experts in agent-based modeling and its applications (expect more leading names):
• Prof. Floor Alkemade (Eindhoven University of Technology)
• Dr. Tommaso Ciarli (UNU-MERIT)
• Prof. Robin Cowan (UNU-MERIT and University of Strasbourg)
• Caterina Croci (ETH Zurich, summer school alumni) TBC
• Francisco Fatas-Villafranca (University of Zaragoza)
• Prof Andre Lorentz (University of Strasbourg)
• Dr. Önder Nomaler (UNU-MERIT)
• Dr. Francesco Pasimeni (Eindhoven University of Technology)
• Anmol Soni (University of Maastricht, summer school alumni)
• Dr. Danilo Spinola (Birmingham City University)
• Dr Serge Stinckwich (UNU-Paris)
• Dr Tania Treibich (University of Maastricht)
• Prof. Marco Valente (University of L’Aquila)
• Prof. Bart Verspagen (UNU-MERIT)
Join us: Application, Participation and Fees
Places are limited to 16 participants to ensure a high-quality learning experience and ample interaction with faculty. We encourage early applications.
Interested applicants should submit a motivation letter (including detailing research interests, how this course aligns with their current or future research goals, programming experience (if any, not a requisite), and career stage) and a CV to ABMschool@merit.unu.edu by July 30th, 2026. Decisions will be communicated by August 7th, 2026. Registration closes on August 17th, 2026.
The course will be held in person at UNU-MERIT, Maastricht (28 September-02 October 2026). Participants should bring their own laptops. We are committed to ensuring accessibility; please contact us to discuss any specific needs.
The participant fee is €400 for students and early-career researchers (postdocs within 2 years of PhD completion), and €550 for others. This covers tuition, course materials, lunch, and coffee/tea breaks. Participants are responsible for their own travel and accommodation expenses. Information on affordable accommodation options in Maastricht will be provided upon acceptance.
Key Dates
- Application deadline: 30 July 2026
- Communication of decision: 7 August 2026
- Registration and fee payment: 17 August 2026
Organising Committee
- Tommaso Ciarli
- Elisa Di Pietro
- Çiğdem Ekiz
- Rebecca Gramiscelli Hasparyk
- Marco Valente
Why UNU-MERIT for your ABM training?
UNU-MERIT is a world-renowned research and training institute focusing on the social, economic, and public policy dimensions of science, technology and innovation; sustainable economic transformations; and human development. Our faculty are at the forefront of applying ABM to complex societal challenges, ensuring you learn from active researchers pushing the boundaries of the field.
Questions?
Please do not hesitate to contact us at ABMschool@merit.unu.edu if you have any queries.
