(D) Now let's run a regression ("help regress"). To keep things simple, use the "robust" option of the regress command to tell Stata to calculate standard errors that are robust to heteroskedasticity. Use the same sample of mothers as in part (A). Create a variable called "Post" that is equal to 1 if the year > 2000 and 0 if the year is less than 1998. Estimate the following regression while \\textbf{excluding the years 1998, 1999, 2000}): Y = g + lPost + gQuebec + 3Post >1: Quebec + 6 (4) Where 6 is the error term, and Y is a binary variable indicating whether the person is employed. Provide the estimates of g, l, g, 3. (E) We can use these four parameter estimates of g, l, g, g to construct the employment rates in each region (Quebec vs nonQuebec) and each period (pre-1998 and post-2000). To get you started, g represents the employment rate in the rest of Canada in the pre1998 period. What is the employment rate in Quebec in the pre- 1998 period? What is the employment rate in Quebec in the post2000 period? What is the employment rate in the rest of Canada in post2000? (F) Which of the four parameter estimates can I call the "causal effect" of the Childcare expansion in Quebec? Based on that, how much did the Childcare expansion increase employment rates among mothers? (G) Estimate the same regression, except for men (again, only men with a youngest child age 5 or less) rather than women. What is the estimated causal effect of the childcare expansion on paternal employment rates