Estimating Causal Effects with Observational Data
PhD School at the Faculty of SCIENCE at University of Copenhagen
This is a toolbox course where 80% of the seats are reserved to PhD students enrolled at the Faculty of SCIENCE at UCPH and 20% of the seats are reserved to PhD students from other Danish Universities/faculties (except CBS). Seats will be allocated on a first-come, first-served basis and according to the applicable rules.
Anyone can apply for the course, but if you are not a PhD student at a Danish university (except CBS), you will be placed on the waiting list until enrollment deadline. After the enrollment deadline, available seats will be allocated to applicants on the waiting list.
Aim and Content
Researchers are often interested in investigating causal relationships, i.e., if and how one variable affects another. While the analysis of causal relationships is ideally done using experimental data, in several research areas (e.g., social sciences) experiments are often infeasible or suffer from important limitations. As a result, most empirical studies in the social sciences and related research areas are based on observational (i.e., nonexperimental) data.
Participants in this course will learn state-of-the-art methods used for investigating causal relationships with observational data. Course participants will also learn how to evaluate and discuss the appropriateness of research designs (“identification strategies”) and empirical methods for analysing causal relationships, and they will learn to choose the most appropriate research designs and empirical methods for a specific research question. This will help participants obtain more credible and reliable results in their own research.
Topics taught in this course include causal directed acyclic graphs (DAGs), methods based on ‘selection-on-observables’, methods based on instrumental variables, synthetic control methods, regression discontinuity designs, difference-in-differences, methods for panel data with staggered treatment, and causal machine learning methods. The course participants will learn the theoretical background and underlying assumptions of these methods as well as how to apply them in real-world analyses.
Learning outcomes
Intended learning outcomes for the students who complete the course:
Knowledge
• Understand causal DAGs.
• Describe methods for causal inference with observational data, including
• ‘Selection-on-observables’
• Instrumental variables
• Synthetic control
• Regression discontinuity designs
• Difference-in-differences
• Methods for panel data with staggered treatment
• Causal machine learning
• Describe the assumptions that need to be fulfilled if the methods listed above should give reliable estimates of causal effects.
Skills
• Construct and interpret causal DAGs and use them to identify causal effects using the DAGitty software.
• Apply methods for causal inference with observational data using (statistical) software such as R, Stata, or Python.
• Assess to which extent assumptions that are required by different causal inference methods with observational data are fulfilled in specific real-world applications.
Competences
• Choose research designs and methods that are appropriate for causal inference with observational data in their research area.
• Critically evaluate the appropriateness of research designs (“identification strategies”) and methods for answering causal questions with observational data in their research area (this refers to their own research, e.g., when discussing strength and weaknesses of causal analyses in their own papers, as well as to the research done by others, e.g., when reviewing manuscripts or assessing the reliability of causal analyses).
Target Group
Ph.D. students at SCIENCE, SUND, SAMF, and other faculties or universities, who aim to investigate causal questions with observational data (e.g., economists, other social scientists, nutritionists, epidemiologists, other health/veterinary scientists, etc.).
Recommended Academic Qualifications
The students should have basic knowledge in statistics (e.g., hypothesis tests, ordinary least-squares (OLS), etc.) obtained, e.g., in the statistics variant of the PhD course "Fundamentals of the PhD education at SCIENCE - module 2" or a similar course.
Research Area
All research areas that apply statistical methods to answer causal research questions with observational data, including economics, other social sciences, nutritional sciences, epidemiology, other health/veterinary sciences, etc.
Teaching and Learning Methods
The course participants are encouraged to read some of the course material before the course starts to be well prepared. The course consists of a combination of lectures and practical exercises. The participants will construct and interpret causal DAGs and they will learn to implement various methods for estimating causal effects. While the teachers will use the R software to present solutions to these exercises, the participants are free to use other software (e.g., Stata or Python). The practical exercises also include group discussions, e.g., about the appropriateness of research designs (“identification strategies”) and empirical methods. The course participants can choose to write a short report (5-10 pages), in which they apply at least one of the methods taught in the course to simulated or real-world observational data, e.g., as a part of their PhD project. Reproducibility of the empirical analysis will play a key role in the lectures, the practical exercises, and in the ‘short report’ (exam).
Type of Assessment
The participants get this course approved with 2.5 ECTS if they attend the lectures, do the practical exercises, and pass a multiple-choice test given at the end of course.
The participants get this course approved with 5 ECTS if they additionally write and submit a short report (see above) that is positively assessed by the teachers. This short reported has to be submitted to the course coordinator no later than 3 months after the end of the course.
Literature
• Angrist, J.D. and Pischke, J.-S. (2009), Mostly Harmless Econometrics, Princeton University Press.
• Angrist, J. D. and Pischke, J. S. (2014). Mastering ‘Metrics: The Path from Cause to Effect. Princeton University Press.
• Bellemare, M.F., Bloem, J.R. and Wexler, N. (2024): The Paper of How: Estimating Treatment Effects Using the Front-Door Criterion. Oxford Bulletin of Economics and Statistics 86: 951-993. https://doi.org/10.1111/obes.12598
• Didelez, V. (2025): Causal Reasoning and Inference in Epidemiology. In Ahrens, W. and Pigeot, I: Handbook of Epidemiology, Springer, New York. https://doi.org/10.1007/978-1-4614-6625-3_74-1
• Digitale, J.C., Martin, J.N. and Glymour, M.M. (2022). Tutorial on directed acyclic graphs. Journal of Clinical Epidemiology, 142, pp.264-267.
• Henningsen, A., Low, G., Wuepper, D., Dalhaus, T., Storm, H., Belay, D. and Hirsch, S (2025): Estimating Causal Effects with Observational Data: Guidelines for Agricultural and Applied Economists. arXiv preprint. https://doi.org/10.48550/arXiv.2508.02310.
• Hernán, M.A. (2018). The C-word: scientific euphemisms do not improve causal inference from observational data. American journal of public health, 108(5), pp.616-619.
• Hernán & Robins (2020): Causal Inference: What If? Chapman & Hall/CRC, Boca Raton (particularly chapters 1, 2, 3, 4, 6, 7, 8 & 16), https://miguelhernan.org/whatifbook
• Huber, M. (2023): Causal analysis: Impact evaluation and Causal Machine Learning with applications in R. MIT Press.
• Huber, M. (2025): Impact Evaluation in Firms and Organizations: With Applications in R and Python. MIT Press.
• Morgan, S.L. and Winship, C. (2014), Counterfactuals and Causal Inference: Methods and Principles for Social Research, 2nd ed. Cambridge University Press.
• Tennant, P.W., Murray, E.J., Arnold, K.F., Berrie, L., Fox, M.P., Gadd, S.C., Harrison, W.J., Keeble, C., Ranker, L.R., Textor, J. and Tomova, G.D. (2021). Use of directed acyclic graphs (DAGs) to identify confounders in applied health research: review and recommendations. International journal of epidemiology, 50(2), pp.620-632.
• Textor, J. (2015), Drawing and analyzing causal DAGs with DAGitty. arXiv preprint arXiv:1508.04633
Course coordinator
Arne Henningsen, Associate Professor, arne@ifro.ku.dk
Teachers
Arne Henningsen, Associate Professor, arne@ifro.ku.dk
Bo Markussen, bomar@math.ku.dk
Christine Winther Bang, cwb@math.ku.dk
Guest Lecturers
When we taught the course in 2025, we invited two renowned experts to give guest lectures on two of the methods covered in this course, respectively. As this worked very well, we plan to include the same or similar guest lectures in the 2026 course as well.
Dates
4-8 May 2026 - 9-17 all days.
Expected frequency
Once per year in teaching block 4.
Course location
UCPH Campus
Course fee
• Participant fee: 0 DKK
• PhD student enrolled at SCIENCE: 0 DKK
• PhD student from Danish PhD school Open market: 0 DKK
• PhD student from Danish PhD school not Open market: 3000 DKK
• PhD student from foreign university: 3000 DKK
• Master's student from Danish university: 0 DKK
• Master's student from foreign university: 3000 DKK
• Non-PhD student employed at a university (e.g., postdocs): 3000 DKK
• Non-PhD student not employed at a university (e.g., from a private company): 8400 DKK
Cancellation policy
• Cancellations made up to two weeks before the course starts are free of charge.
• Cancellations made less than two weeks before the course starts will be charged a fee of DKK 3.000
• Participants with less than 80% attendance cannot pass the course and will be charged a fee of DKK 5.000
• No-show will result in a fee of DKK 5.000
• Participants who fail to hand in any mandatory exams or assignments cannot pass the course and will be charged a fee of DKK 5.000
Course fee and participant fee
PhD courses offered at the Faculty of SCIENCE have course fees corresponding to different participant types.
In addition to the course fee, there might also be a participant fee.
If the course has a participant fee, this will apply to all participants regardless of participant
type - and in addition to the course fee.