Online lecture 8 – Session 1: OLS regression models and causal inference

Lecturers: Dr. Auke Rijpma, Research Institute for History and Art History, Utrecht University

Content:Historical demographers often rely on survival models to analyze events like mortality and health outcomes. While these are powerful and appropriate tools, the more basic OLS (Ordinary Least Squares) model is a very interesting and often robust alternative, especially when data doesn’t perfectly adhere to the requirements of more complex models. What’s more, there is a well-developed literature on how to perform causal inference with OLS models, allowing researchers to go beyond simple correlation to identify meaningful causal relationships. This lecture will introduce the basic OLS model and then cover a number of key causal inference strategies. We will also demonstrate practical estimation procedures using the powerful fixest package in R, highlighting how these methods can be applied to historical demographic data.

Objectives:

• Discuss the basics of the OLS model
• Introduce important causal inference strategies that can be performed with extensions of the OLS model
• Demonstrate OLS estimation in R’s fixest library

Requirements: Active participation.

Recommended reading:

• Angrist, Joshua D., and Jörn-Steffen Pischke. 2015. Mastering ’metrics: The Path from Cause to Effect. Princeton University Press.
• Zwart, Pim de, Daniel Gallardo-Albarrán, and Auke Rijpma. 2022. ‘The Demographic Effects of Colonialism: Forced Labor and Mortality in Java, 1834–1879’. The Journal of Economic History 82 (1): 211–49. https://doi.org/10.1017/S0022050721000577.

Slides for Online lecture 8 – Session 1- OLS regression models and causal inference