uppose that you want to analyse the determinants of the murder rate in the United Kingdom using the following model: Murderrt = β0 + β1Convict + β2Unempt + β3Ymalet + ut where Murderrt is the number of murders per 100,000 persons during year t, Convic is the murder conviction rate, Unemp is the national unemployment rate, and Ymale is the fraction of males aged between 18 and 25. Data for all the variables are available for 25 years. a) You estimate this model using the OLS estimator. Explain the properties of the OLS estimator, β 1 ^, and the estimator of the variance, Var(β 1 ^), if the error term ut is serially correlated. b) You suspect the presence of the AR (1) serial correlation in the error term ut and decided to test for it using the Breusch-Godfrey test. State the null hypothesis for the Breusch-Godfrey test. c) Eviews gave you the F-statistic of 8.21 for the Breusch-Godfrey test. Do you reject the null hypothesis? Explain why or why not. If you did find evidence of serial correlation, how would you deal with it?
Suppose that you want to analyse the determinants of the murder rate in the United Kingdom using the following model:
Murderrt = β0 + β1Convict + β2Unempt + β3Ymalet + ut
where Murderrt is the number of murders per 100,000 persons during year t, Convic is the murder conviction rate, Unemp is the national unemployment rate, and Ymale is the fraction of males aged between 18 and 25. Data for all the variables are available for 25 years.
a) You estimate this model using the OLS estimator. Explain the properties of the OLS estimator, β 1 ^, and the estimator of the variance, Var(β 1 ^), if the error term ut is serially correlated.
b) You suspect the presence of the AR (1) serial
c) Eviews gave you the F-statistic of 8.21 for the Breusch-Godfrey test. Do you reject the null hypothesis? Explain why or why not. If you did find evidence of serial correlation, how would you deal with it?
Step by step
Solved in 3 steps with 3 images
