In this video, I want to provide an introduction to simultaneous equations in econometrics. I'm going to explain some of the issues that arise when you try to estimate simultaneous equation models. What do I mean by a simultaneous equation model? Well, an example would be if we thought an individual's wage was determined by their level of status within a society. So, only those of a high status tend to be given jobs that pay more. Furthermore, we think that another determinant of wage might be some sort of governmental policy, which I'm going to represent here by this policy variable. There are also a range of other factors which are contained in this error term, epsilon 1. Furthermore, we think that status is itself determined by the level of wage which an individual earns because there is some sort of feedback between these two things. Furthermore, there is also a status gain for being married or having a family, which I'm representing with the term M. There are a range of other factors which also affect an individual's status within society, which are contained within this error term, epsilon 2. So, we've got two equations and there is obviously some coupling between these two equations because wage in the second equation appears in the independent variable where it is the dependent variable of the first equation, and similarly, status appears as the dependent variable of the second equation but it's actually the independent variable from the first. So we can draw arrays like this, indicating the interaction between the two equations. These two equations together form what we call a simultaneous equation model, sometimes abbreviated as Shem. Now, what's the problem with estimating these two equations using OLS? Well, the issue is that we actually get something known...