Suppose you want to estimate the effect of Philadelphia's sugar tax, which imposes a 1.5 cent...

Transcribed Image Text:

Suppose you want to estimate the effect of Philadelphia's sugar tax, which imposes a 1.5 cent per ounce tax on sugar sweetened beverages. To do so, you collect data on monthly purchases of sugar sweetened beverages (in oz.) for the cities of Philadelphia and Pittsburgh, the latter of which does not have a sugar tax. The following table presents average purchases of sugar sweetened beverages. Philadelphia (Treatment Group) 15.75 16.10 Pre-Tax Post-Tax Pittsburgh (Control Group) 17.25 18.50 1. By how much do monthly sugar sweetened beverage purchases change over time in Philadelphia? 2. By how much do monthly sugar sweetened beverage purchases change over time in Pittsburgh? 3. Using your answers to (1.) and (2.), what is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? Hint: Subtract your answer to (2.) from your answer to (1.). 5. Is your answer to (4.) the same as your answer to (3.)? 4. You decide to estimate a simple regression model that does not include any control variables. Using the following regression line, purchasest = 17.25 -1.50treatment; +1.25post - 0.90treatment * postit, what is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? 6. Instead, you decide to estimate a model with control variables (including age, level of education, and income). The new regression line is purchasest = 10.50 -1.15treatment₁ + 0.50post 0.15treatment * postit. What is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? 7. Based on your answers to (3.), (4.), and (6.), does the sugar tax decrease monthly purchases of sugar sweetened beverages? 8. Assuming you observe monthly purchases by the same individuals over time, would you present results with conventional, clustered, or Newey-West standard errors? 9. The graph below plots average monthly sugar sweetened beverage purchases for Philadelphia and Pittsburgh. Assuming that the sugar tax is imposed at time 0, can we use a difference-in-differences model to estimate the effect of the sugar tax on monthly sugar sweetened beverage purchases (i.e., are the growth rates in purchases the same in Philadelphia and Pittsburgh prior to the implementation of the sugar tax)? 24 Sugar Sweetened Beverage Purchases m -3 -2 -1 0 Time 1 -Philadelphia -Pittsburgh 2 3 Suppose you want to estimate the effect of Philadelphia's sugar tax, which imposes a 1.5 cent per ounce tax on sugar sweetened beverages. To do so, you collect data on monthly purchases of sugar sweetened beverages (in oz.) for the cities of Philadelphia and Pittsburgh, the latter of which does not have a sugar tax. The following table presents average purchases of sugar sweetened beverages. Philadelphia (Treatment Group) 15.75 16.10 Pre-Tax Post-Tax Pittsburgh (Control Group) 17.25 18.50 1. By how much do monthly sugar sweetened beverage purchases change over time in Philadelphia? 2. By how much do monthly sugar sweetened beverage purchases change over time in Pittsburgh? 3. Using your answers to (1.) and (2.), what is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? Hint: Subtract your answer to (2.) from your answer to (1.). 5. Is your answer to (4.) the same as your answer to (3.)? 4. You decide to estimate a simple regression model that does not include any control variables. Using the following regression line, purchasest = 17.25 -1.50treatment; +1.25post - 0.90treatment * postit, what is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? 6. Instead, you decide to estimate a model with control variables (including age, level of education, and income). The new regression line is purchasest = 10.50 -1.15treatment₁ + 0.50post 0.15treatment * postit. What is the difference-in-differences estimate of the effect of the sugar tax on monthly sugar sweetened beverage purchase? 7. Based on your answers to (3.), (4.), and (6.), does the sugar tax decrease monthly purchases of sugar sweetened beverages? 8. Assuming you observe monthly purchases by the same individuals over time, would you present results with conventional, clustered, or Newey-West standard errors? 9. The graph below plots average monthly sugar sweetened beverage purchases for Philadelphia and Pittsburgh. Assuming that the sugar tax is imposed at time 0, can we use a difference-in-differences model to estimate the effect of the sugar tax on monthly sugar sweetened beverage purchases (i.e., are the growth rates in purchases the same in Philadelphia and Pittsburgh prior to the implementation of the sugar tax)? 24 Sugar Sweetened Beverage Purchases m -3 -2 -1 0 Time 1 -Philadelphia -Pittsburgh 2 3

Answer rating: 100% (QA)

1 In Philadelphia monthly sugar sweetened beverage purchases change over time as follows Pretax 1575 oz Posttax 1610 oz The change in monthly purchase... View the full answer