Question 1 (25 points ) You are regressing quarterly public health care expenditures in constant dollars in Canada on the proportion of the population aged 65 years old and above and on the number of drug and medical devices patents granted in that quarter. (3) Does it make sense to include a lagged dependent variable as a re- gressor? Why? What concerns would you have if you do this? (4) (b) Should you control for trends? Why or why not? How would you do so? (4) (c) Should you control for seasonality? Why or why not? How would you do so? (4) (d) Should you test for serial correlation of the errors? How ( State 1 test for this and describe how you would perform it)? What is it? What are its consequences? (5) (e) If you were to find that your errors are serially correlated, how could you correct your model for it? (5) (f) Explain the difference between an AR and an MA process. (3)