This chapter discusses the concept of Granger causality for panel datasets that are both wide (many individuals) and long (many time periods). Such datasets are obtained for instance when sets of countries are observed over the long run. In such settings, classical issues of time series econometrics, such as (non-)stationarity and (non-)causality, also arise. After a brief theoretical presentation of Granger causality, we provide a complete empirical application using publicly available data from the World Bank and Climate Watch. We investigate Granger causality between economic growth and CO emissions for a set of 130 countries observed over 41 years. The empirical exercise is conducted with the statistical software Stata and is fully reproducible; thanks to the code that is provided. All steps of the analysis are carefully explained and the results are intuitively presented.