What are control variables good for and why do we use them? How can we use control variables to solve endogeneity problems?