Panel Data Analysis Using Stata


If you have repeated observations of voters, countries, companies, or other units of interest that vary over time, then you have panel data. Variation over time gives us more insight than a cross-section, which only provides a snapshot at one moment in time. With panel data we can model the heterogeneity of our unit of interest and how it changes over time.

In this course we will learn how to analyze panel data using Stata. As such, the course will discuss the theory of panel data and at the same time illustrate how to use panel data estimators. We will cover linear and non-linear panel data models, which are a natural extension from their cross-sectional counterparts. We will also discuss dynamic estimators that explicitly model the dynamic structure of panel data, the variation of the individuals or other units of interest, over time.

Below is a list of the topics we will cover:

  • A quick introduction to Stata
  • Overview of basic regression analysis
  • Estimators for linear models with random effects
  • Fixed-effect estimators for linear models
  • Instrumental-variables estimators
  • Models with endogenous variables
  • Dynamic panel-data models
  • The Arellano–Bond estimator
  • The Arellano–Bover/Blundell–Bond estimator
  • Random- and fixed-effects estimators for binary models
  • Random- and fixed-effects estimators for count-data models

Prerequisites: You must be familiar and comfortable with the mathematical material assumed in an advanced undergraduate econometrics class such as the material in appendices A--E in Introductory Econometrics, 5th edition by Jeffrey Wooldridge.

Fee: Members = $1500; Non-members = $2800

Tags: panel data, Stata

Course Sections