Endogeneity occurs where an explanatory variable is present within your regression model which is correlated to the error term. This is therefore referred to as an endogenous variable. This violates Classical Assumption number 3 which states that there is no correlation between any of the explanatory variables and the error term.
Broad assumptions regarding an independent variable can often result in endogeneity in a linear regression model. For example, let’s say that we are running a regression to assess the relationship between age and earnings. Now we would assume that as someone ages that they begin to earn more money. However, if we assume that age is the only determinant of earnings then this variable is going to be correlated to the error term.
Because there are other regressors know as instrument variables (IV) such as experience and education for example which will play a key role in determining the level of income that an individual will attain. Therefore our explanatory variable of age will be correlated to variables accounted for within the error term.
As discussed in a previous post on the error term, the error term is the difference between what your model estimates as Y and the true population value of Y. Now if you have not accounted for the inclusion of certain variables in your models, then these will be correlated with your regressor and will be accounted for within the error term.
If endogeneity is present in your regression, this model’s estimates cannot be relied upon.
Find us on facebook:
Check out our World-Class Econometrics courses here: