Consider the following model of the Philips curve. In this model inflation is regressed upon unemployment using a sample of 56 observations, where INF denotes a measure of the inflation rate and UNEM denotes a measure of the unemployment rate. An initial analysis of this model produced the results reported in Tables Two and Three. Table Two Dependent Variable: INF Method: Least Squares Sample: 1948 2003 Included observations: 56 Variable Coefficient Std. Error t-Statistic Prob. C UNEM 1.053565 1.547957 0.680617 0.4990 0.502378 0.265562 1.891752 0.0639 Akaike info criterion 5.051084 Schwarz criterion 5.123418 Table Three Breusch-Godfrey Serial Correlation LM Test F-statistic Obs*R-squared 31.52897 Prob. F(1,53) 20.88778 Prob. Chi-Square(1) 0.0000 0.0000 Further analysis produced the results in Tables Four and Five. Table Four Dependent Variable: INF Method: Least Squares Sample (adjusted): 1950 2003 Included observations: 54 after adjustments Variable INF(-1) UNEM UNEM(-1) UNEM(-2) Coefficient Std. Error t-Statistic Prob. 1.409911 1.139949 1.236820 0.2220 0.833123 0.104146 7.999534 0.0000 -0.421441 0.314574 -1.339722 0.1865 0.359092 -0.567296 0.5731 0.258068 1.953120 0.0565 -0.203712 0.504038 Schwarz criterion 4.362220 Akaike info criterion 4.178055 Table Five Breusch-Godfrey Serial Correlation LM Test: F-statistic Obs*R-squared 2.325239 Prob. F(1,48) 2.495029 Prob. Chi-Square(1) 0.1339 0.1142 a) Explain what is meant by serial correlation and discuss its underlying causes. b) Report the findings from serial correlation testing presented in Tables Three and Five. Compare with the Durbin-Watson test, explain in detail the advantages and the shortcomings of the Breusch-Godfrey Serial Correlation LM Test. c) Compare the two models presented in Tables Two and Four on the basis of the calculated values of the information criteria considered. Using appropriate justification, explain which model you prefer. d) A new ARDL (2,3) model of INF is estimated, with no serial autocorrelation detected for this model. The calculated values of the Schwarz Information Criterion and Akaike Information criterion for this model are given as: The Schwarz Information Criterion = 4.443694 The Akaike Information Criterion = 4.183467 Compare this model with those presented in Tables Two and Four. Which of the three models do you prefer? Explain your