the question

What is a reason we might need to include control variables when we use IV estimation? a. We want to include regressors in order reduce the standard error of the causal parameter. b. We need to capture that treatment is assigned with different probabilities for different values of control variables. c. We like to include regressors in order to improve the fit of the model. d. Including control variables makes the first stage stronger