1. Are there pre-treatment differences between control cities and treatment cities? How can you tell? 

2. Is the coefficient on treatment in model 2 an unbiased estimate of the effect of the carbon tax treatment on greenhouse gas emissions? Why or why not? 

3. Interpret the three coefficients in model 3.

4. What assumption is necessary for this differences-in-differences framework to provide unbiased estimates of the effect of the carbon tax on greenhouse gas emissions? 

Transcribed Image Text:

The following table provides results from a differences-in-differences regression to estimate the effects of a carbon tax on greenhouse gas emissions. The regressions were carried out on a dataset which provides annual greenhouse gas emission for several cities from the years 2010-2017. Treatment is defined as the introduction of a carbon tax in the year 2013. Cities where the carbon tax was introduced are part of the treatment group, and cities where the carbon tax was not introduced are part of the control group. Greenhouse gas emissions are measured in mega tonnes. The title of the column refers to the dependent variable in that regression. [9 marks] Model 1 Model 2 Model 3 Annual Dependent Greenhouse gas Greenhouse gas greenhouse gas Variable: emission in 2010 emission in 2017 emission Treatment -122.34*** -236.14*** -119.72*** (32.43) (62.54) (23.46) Post -17.82*** (4.56) Treatment X Post -104.67*** (24.51) Adjusted R-squared 0.216 0.18 0.42 Observations 60 60 480 ** p<0.10, p<0.05, *** p<0.01 The following table provides results from a differences-in-differences regression to estimate the effects of a carbon tax on greenhouse gas emissions. The regressions were carried out on a dataset which provides annual greenhouse gas emission for several cities from the years 2010-2017. Treatment is defined as the introduction of a carbon tax in the year 2013. Cities where the carbon tax was introduced are part of the treatment group, and cities where the carbon tax was not introduced are part of the control group. Greenhouse gas emissions are measured in mega tonnes. The title of the column refers to the dependent variable in that regression. [9 marks] Model 1 Model 2 Model 3 Annual Dependent Greenhouse gas Greenhouse gas greenhouse gas Variable: emission in 2010 emission in 2017 emission Treatment -122.34*** -236.14*** -119.72*** (32.43) (62.54) (23.46) Post -17.82*** (4.56) Treatment X Post -104.67*** (24.51) Adjusted R-squared 0.216 0.18 0.42 Observations 60 60 480 ** p<0.10, p<0.05, *** p<0.01

Answer rating: 100% (QA)

Solution 1 Treatment in 2010 refers to control cities and Treatment in 2017 refers to treatment citi... View the full answer