3. For the following questions, use the _ ' -- ' data located within this -_ module on Games. For this question, you are interested in examining a policy implemented in Michigan in 2005 that provides extra crop insurance for hay (remember from 1.d. that growing hay improves water quality!) to induce more farmers to grow hay. a. Perform a time-series analysis of the \"pre\" and \"post" periods for the Michigan policy (a difference of means test). Did the policy result in a statistical difference in means of acres of hay farmed in Michigan before and after the policy? b. What are some potential problems with your analysis in a? Do your results identify a causal relationship between policy and outcome? c. Assuming Wisconsin is a reasonable control sample [I am not asking you to assess pre- trends in the outcome), what is the value of the differencoin-difference estimator? What does this value represent? Does this represent a causal relationship between policy and outcome? d. Would you recommend implementing this hay insurance policy in Wisconsin? Why or why not?