In this module, “panel data” refers to the pooling of observations on a cross-section of countries, households, firms, etc. over a number of time periods. We use panel data techniques for example to control for individual heterogeneity or to study the dynamics of adjustment. Panel data allows for more informative results, more variability, more degrees of freedom and more efficiency. This module focuses on statistical theory and empirical estimation, interpretation and evaluation of economic relationships, within a panel data context. The module covers both techniques applicable to stationary and non-stationary panel data sets, and begins with an introduction to one-way error component models (either including individual-specific or period-specific (time) effects), followed by two-way error component models (including individual-specific and time effects simultaneously). Estimation techniques include fixed effects (LSDV or “Within” estimation) and random effects estimation. Hypothesis testing includes tests for poolability (pooled vs. individual regressions), tests for fixed effects, random effects, and specification (exogeneity of the X-regressors). It also includes various tests for serial correlation and heteroscedasticity and the correction thereof. The section on stationary panel data techniques concludes with a discussion of seemingly unrelated regression (SUR) models. In the non-stationary panel data section we discuss unit root testing in the panel context, estimation of non-stationary panels and tests for co integration.