1. Consider the following regression: \text { InGDPpC }=\beta_{0}+\beta_{1} \text { Institutions } 1+u_{1} where the dependent variable is In of GDP per capita, the explanatory variable is a measure of institutional quality (a higher value implies better quality institutions),and the subscript i represents countries. [7 marks] a) Draw a scatter plot that demonstrates how this regression would be biased and explain how your scatter plot demonstrates the bias. For simplicity, assume that there are no other sources of bias when creating your scatter plot. Your scatterplot should be clearly labelled and easy to understand. [2 marks] b) Suppose you estimate the above equation using OLS and the estimated value ofB₁ is 0.23. Interpret this coefficient. [1 mark] c) Now, suppose you have a valid instrumental variable. Do you expect the TSL Sestimate to be greater than, less than, or the same as the OLS estimate of B₁?Explain your answer. [2 marks] d) Explain whether or not panel data would be useful for addressing simultaneous causality bias in this context. [2 marks]

Fig: 1

Fig: 2

Fig: 3

Fig: 4

Fig: 5

Fig: 6

Fig: 7