Please use the Fatalities dataset from the AER package in R to answer this question. Please carefully read the document of data description. Government wants to reduce traffic fatalities by increasing the tax on alcoholic beverages. Thus, they need help understanding the relationship between the traffic fatality rate and the alcohol taxes. The bassic model is: FatalityRateit=0+1BteerTaxit+it (a) Suppose that the fatality rate FatalityRate is defined as the number of vehicle fatalities divided by population. Please generate FatalityRate and estimate the basic model (i.e. a model without any fixed effects). Please interpret all the coefficient estimates in your regression table (b) Determine bow many years and different states there are in the data set. Is it a balanced panel? Explain. (Hints: A panel dataset is considered "balanced" when all entities are observed for all time periods. If there are any missing values, it is called an unbalanced panel.) (c) Now add a full set of state dummies into the basic model (denoted as model 2). Does the coefficient and significance level of BeerTax it change? Please explain, using your own language, why there is a change (or lack of it). (d) Now add a full set of year dummies into the basic model (denoted as model 3). Does the coefficient and significance level of BeerTaxit change? Please explain, using your own language, why there is a change (or lack of it). (e) Now use the package plm in R to estimate a model with state fixed effects (denoted as model 4). Compare the coefficients in model 2 and model 4 . Are the results identical? Why or why not? (f) Now use the package plm in R to estimate the basic model with year-fixed effects (denoted as model 5). Compare the coefficients in model 3 and model 5 . Are the results identical? Why or why not? (g) A scholar argues that the state of the nation's economy will affect the quality of road infrastructure and thus influence traffic fatalities. Supposing we have the data of GDP (Groes Domestic Product) for every year, do we need to add that variable into model 3 to control for the national economic condition? Why or why not? (h) Follow the steps for a state fixed effects regression via demeaning and estimate a new model on the entity-demeaned variables. Are the results identical to model 4? Are they identical to model 2? (i) Follow the steps for a year fixed effects regression via demeaning and estimate a model on the time-demeaned variables. Are the results identical to model 5? Are they identical to model 3 ? Please use the Fatalities dataset from the AER package in R to answer this question. Please carefully read the document of data description. Government wants to reduce traffic fatalities by increasing the tax on alcoholic beverages. Thus, they need help understanding the relationship between the traffic fatality rate and the alcohol taxes. The bassic model is: FatalityRateit=0+1BteerTaxit+it (a) Suppose that the fatality rate FatalityRate is defined as the number of vehicle fatalities divided by population. Please generate FatalityRate and estimate the basic model (i.e. a model without any fixed effects). Please interpret all the coefficient estimates in your regression table (b) Determine bow many years and different states there are in the data set. Is it a balanced panel? Explain. (Hints: A panel dataset is considered "balanced" when all entities are observed for all time periods. If there are any missing values, it is called an unbalanced panel.) (c) Now add a full set of state dummies into the basic model (denoted as model 2). Does the coefficient and significance level of BeerTax it change? Please explain, using your own language, why there is a change (or lack of it). (d) Now add a full set of year dummies into the basic model (denoted as model 3). Does the coefficient and significance level of BeerTaxit change? Please explain, using your own language, why there is a change (or lack of it). (e) Now use the package plm in R to estimate a model with state fixed effects (denoted as model 4). Compare the coefficients in model 2 and model 4 . Are the results identical? Why or why not? (f) Now use the package plm in R to estimate the basic model with year-fixed effects (denoted as model 5). Compare the coefficients in model 3 and model 5 . Are the results identical? Why or why not? (g) A scholar argues that the state of the nation's economy will affect the quality of road infrastructure and thus influence traffic fatalities. Supposing we have the data of GDP (Groes Domestic Product) for every year, do we need to add that variable into model 3 to control for the national economic condition? Why or why not? (h) Follow the steps for a state fixed effects regression via demeaning and estimate a new model on the entity-demeaned variables. Are the results identical to model 4? Are they identical to model 2? (i) Follow the steps for a year fixed effects regression via demeaning and estimate a model on the time-demeaned variables. Are the results identical to model 5? Are they identical to model 3
