* Reference: Chapters 10 and 11 of * Jeffrey M. Wooldridge, Introductory Econometrics: A Modern Approach, * South-Western College Publishing, 2000. SAMPLE 1 72 READ (FERTIL3.shd) gfr pe year t tsq pe_1 pe_2 pe_3 pe_4 & pill ww2 tcu cgfr cpe cpe_1 cpe_2 cpe_3 cpe_4 gfr_1 & cgfr_1 cgfr_2 cgfr_3 cgfr_4 gfr_2 SET MISSVALU=-999 SET SKIPMISS * Example 10.4 * Equation (10.18), p. 325. OLS gfr pe ww2 pill * Equation (10.19), p. 326. * Incorporate a distributed lag (see the chapter DISTRIBUTED-LAG MODELS * in the SHAZAM User's Reference Manual). OLS gfr pe(0.2) ww2 pill * The SHAZAM estimation results report: * TESTS ON LAGGED COEFFICIENTS * VARIABLE SUM(COEFS) STD ERROR T-RATIO P-VALUE | JOINT-F P-VALUE * PE 0.10072 0.298E-01 3.38 0.001 | 3.97 0.012 * The JOINT-F gives the F-test for testing the joint significance * of pe(t), pe(t-1) and pe(t-2). * The SUM(COEFS) reports the sum of the lag coefficients that * gives the estimate of the long-run propensity (LRP). * The STD ERROR is the standard error of this estimate. * Example 10.8 * Equation (10.34), p. 336. * Generate a time trend GENR t=TIME(0) OLS gfr pe ww2 pill t * Equation (10.35), p. 336. * Introduce a quadratic trend GENR tsq=t**2 OLS gfr pe ww2 pill t tsq * Example 11.6 * Estimate an AR(1) model - see Example 11.3, p. 354. * The lag notation is described in the SHAZAM User's Reference * Manual in the chapter DISTRIBUTED LAG MODELS. OLS gfr gfr(1.1) OLS pe pe(1.1) * The estimated coefficient on the regressor is the * estimate of the first order autocorrelation coefficient. * Now estimate using first differences. * Equation (11.26), p. 365. * Adjust the sample period for the undefined first observation. SAMPLE 2 72 OLS cgfr cpe * Equation (11.27), p. 365. OLS cgfr cpe(0.2) STOP