ABSTRACT
Section IV of this book introduces both panel data and instrumental variables. Panel data has an element of magic about it; it allows us to observe the unobservable. Unobservable effects are one of the most important reasons why it is difficult to argue for causality. We always face the objection that there is something unobserved driving the outcome we observe. Clearly, being able to control for that unobservable would be a big step forward. In this chapter we show how we can use panel data to control for time-invariant unobservables (a concept that we will explain shortly). If the unobservables correlated with our explanatory variables do vary over time, then we need to take another step. In this case we need a variable which can shift the explanatory variable but not otherwise affect our dependent variable. Such variables are called instruments, and we introduce them in Chapter 11.
