Assignment #2

ECON 323: Econometric Analysis 2 – Winter 2019

Due February 14th, 2019, in class at 11:30am

Instructions: While cooperating on the assignment is encouraged, pla- giarism is not. I will only accept hand written assignment submitted in person. DO NOT SUBMIT YOUR ASSIGNMENT ELECTRONICALLY.

No late assignment will be accepted; late assignments (after 8:30am on the due date) will receive a mark of zero. Show your work as no marks will be allocated for the final answer alone.

Use Stata or R to do these. If you choose to use another software, please get my approval as soon as possible.

Question 1

Use Canada’s Alcohol and Other Drugs Survey, 1994 [Canada] to do the assignment (which you can access on ODESI).

You are trying to estimate a model of whether people are in the labour force. You believe that this is a function of province of residence, the number of cigarettes smoked and how many times the individual binge drank in the past year. You limit your sample to individuals who have smoked before.

a) Run the above model using OLS. Report your findings.

b) You think that there may be endogeneity between binge drinking and whether the individual is part of the labour force. Using whether the individual’s father has a drinking problem a, run the regression using the built-in command in Stata and program it yourself. Do you get the same results?

c) Test whether there is a difference between your results in (b) and in (a). Were you justified to think that endogeneity was present?

d) Is the variable in (b) a “good instruments”. Why or why not? I expect both a theoretical and a quantitative answer here.

e) Regardless of your answer in (f) suggest a different instrument. Ex- plain why your instrument would be better. If your variable actually is present in the dataset, you will earn an extra 2 (bonus) marks.