Data Analysis for Business, Economics, and Policy
A complete course in data analysis by Gábor Békés and Gábor Kézdi: data wrangling, regression, prediction with machine learning, and causal analysis — taught through 47 case studies using real-world data, with all code in R, Python, and Stata.
Find your path
Instructors
Adopt and teach with the book: slides for every chapter, teaching guide, solutions, and adoption examples.
Students
Learn with the book: quick links, coding setup in R, Python, or Stata, practice Q&A, and study advice.
Data & Code
Everything is reproducible: raw and clean datasets on OSF, full code for every case study on GitHub.
Data Analysis with AI
New: teaching and doing data analysis in the age of LLMs — a full course and materials in progress.
What the book covers
A complete, curated curriculum that equips future data analysts with the most important tools, methods, and skills — through the entire process of data analysis, to answer real-life questions.
I · Data Exploration
Data collection and quality, tidy data and wrangling, exploratory analysis and visualization, generalizing from data, hypothesis testing.
II · Regression Analysis
Non-parametric and linear models, functional form, internal and external validity, probability models, time series regressions.
III · Prediction
Loss functions, cross-validation, LASSO, tree-based machine learning (CART, random forest, boosting), classification, forecasting.
IV · Causal Analysis
Potential outcomes and DAGs, experiments, matching, difference-in-differences, panel data methods, synthetic control, event studies.
Case studies: global and diverse
Each of the 47 case studies begins with a real question and ends with an answer, based on real data and the methods taught in that chapter. For example:
- Estimating gender and age differences in earnings (USA). More
- Management quality, firm size and family ownership (Mexico, International). More
- Predicting company default with machine learning (EU). More
- Working from home and employee performance (China). More
- Identifying successful football managers, and the effect of a change (UK). More
Endorsements
Comprehensive and accessible… exactly what is needed.
David Card UC Berkeley · Nobel laureate
A beautiful integration of Econometrics and Data Science.
Joshua Angrist MIT · Nobel laureate
Must purchase for anyone doing applied work… Perfect for data scientists of all stripes (including Econ).
Scott Cunningham Baylor University · Author of Causal Inference: The Mixtape
Adopted by 90+ programmes worldwide
In Economics, Finance, Analytics, Business, and Public Policy — from Columbia and Michigan to Bocconi, CEU, and beyond. Full list of courses →
About the authors
Gábor Békés and Gábor Kézdi at Balatonudvari, Hungary (July 2018). Photo by Anna Fetter.
We could not have done this alone. Far from it. So, we are grateful, really.