---------------------------------------------------------------------------------- name: log: /disk/oldadmin/homes/web/html/stata/tsls_test.log log type: text opened on: 21 May 2013, 16:43:21 . timer clear . rename ln_wage y1 . rename wks_work y2 . rename hours z1 . rename ttl_exp x1 . rename age x2 . gen z2 = uniform() * 10 . gen y3 = uniform() * 20 . gen z3 = uniform() * 30 . . xtset idcode panel variable: idcode (unbalanced) . timer on 1 . xtivreg2 y1 (y2 = z1) z3 x1 x2 , fe first cluster(idcode) Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 First-stage regressions ----------------------- First-stage regression of y2: FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 OLS estimation -------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 2865.49 Prob > F = 0.0000 Total (centered) SS = 15931490.53 Centered R2 = 0.3229 Total (uncentered) SS = 15931490.53 Uncentered R2 = 0.3229 Residual SS = 10787093.38 Root MSE = 21.63 ------------------------------------------------------------------------------ | Robust y2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- z3 | -.0106124 .0164473 -0.65 0.519 -.042858 .0216333 x1 | 2.272067 .0998863 22.75 0.000 2.076236 2.467898 x2 | 1.121418 .074994 14.95 0.000 .9743888 1.268446 z1 | .2723216 .0197628 13.78 0.000 .2335759 .3110674 ------------------------------------------------------------------------------ Included instruments: z3 x1 x2 z1 ------------------------------------------------------------------------------ F test of excluded instruments: F( 1, 4114) = 189.87 Prob > F = 0.0000 Angrist-Pischke multivariate F test of excluded instruments: F( 1, 4114) = 189.87 Prob > F = 0.0000 Summary results for first-stage regressions ------------------------------------------- (Underid) (Weak id) Variable | F( 1, 4114) P-val | AP Chi-sq( 1) P-val | AP F( 1, 4114) y2 | 189.87 0.0000 | 189.94 0.0000 | 189.87 NB: first-stage test statistics cluster-robust Stock-Yogo weak ID test critical values for single endogenous regressor: 10% maximal IV size 16.38 15% maximal IV size 8.96 20% maximal IV size 6.66 25% maximal IV size 5.53 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. Underidentification test Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified) Ha: matrix has rank=K1 (identified) Kleibergen-Paap rk LM statistic Chi-sq(1)=180.46 P-val=0.0000 Weak identification test Ho: equation is weakly identified Cragg-Donald Wald F statistic 239.14 Kleibergen-Paap Wald rk F statistic 189.87 Stock-Yogo weak ID test critical values for K1=1 and L1=1: 10% maximal IV size 16.38 15% maximal IV size 8.96 20% maximal IV size 6.66 25% maximal IV size 5.53 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and orthogonality conditions are valid Anderson-Rubin Wald test F(1,4114)= 0.24 P-val=0.6226 Anderson-Rubin Wald test Chi-sq(1)= 0.24 P-val=0.6225 Stock-Wright LM S statistic Chi-sq(1)= 0.24 P-val=0.6233 NB: Underidentification, weak identification and weak-identification-robust test statistics cluster-robust Number of clusters N_clust = 4115 Number of observations N = 27177 Number of regressors K = 4 Number of endogenous regressors K1 = 1 Number of instruments L = 4 Number of excluded instruments L1 = 1 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 345.71 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = 0.1258 Total (uncentered) SS = 2344.706896 Uncentered R2 = 0.1258 Residual SS = 2049.727584 Root MSE = .2981 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008881 .0018214 -0.49 0.626 -.004458 .0026818 z3 | -.0001671 .0002293 -0.73 0.466 -.0006165 .0002823 x1 | .0406883 .004789 8.50 0.000 .0313021 .0500746 x2 | -.0048113 .0022836 -2.11 0.035 -.0092871 -.0003355 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 180.457 Chi-sq(1) P-val = 0.0000 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 239.138 (Kleibergen-Paap rk Wald F statistic): 189.875 Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38 15% maximal IV size 8.96 20% maximal IV size 6.66 25% maximal IV size 5.53 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 Included instruments: z3 x1 x2 Excluded instruments: z1 ------------------------------------------------------------------------------ . timer off 1 . . preserve . timer on 2 . tsls y1 (y2 = z1) z3 x1 x2, demean fe(idcode) first cluster(idcode) replace Linear regression Number of obs = 27177 F( 4, 4114) = 2865.38 Prob > F = 0.0000 R-squared = 0.3229 Root MSE = 19.925 (Std. Err. adjusted for 4115 clusters in idcode) ------------------------------------------------------------------------------ | Robust y2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- z1 | .2723216 .0197632 13.78 0.000 .2335751 .3110682 z3 | -.0106124 .0164476 -0.65 0.519 -.0428586 .0216339 x1 | 2.272067 .0998882 22.75 0.000 2.076232 2.467902 x2 | 1.121418 .0749954 14.95 0.000 .9743861 1.268449 _cons | 5.98e-08 2.84e-08 2.11 0.035 4.16e-09 1.15e-07 ------------------------------------------------------------------------------ ( 1) z1 = 0 F( 1, 4114) = 189.87 Prob > F = 0.0000 tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = 0.1258 Total (uncentered) SS = 2344.706897 Uncentered R2 = 0.1258 Residual SS = 2049.727585 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008881 .0018214 -0.49 0.626 -.004458 .0026818 z3 | -.0001671 .0002293 -0.73 0.466 -.0006165 .0002823 x1 | .0406883 .004789 8.50 0.000 .0313021 .0500746 x2 | -.0048113 .0022836 -2.11 0.035 -.0092871 -.0003355 _cons | -6.34e-10 7.71e-10 -0.82 0.411 -2.14e-09 8.77e-10 ------------------------------------------------------------------------------ . timer off 2 . . . timer on 3 . tsls y1 (y2 = z1) z3 x1 x2, fe(idcode) first cluster(idcode) Linear regression Number of obs = 27177 F( 4, 4114) = 2865.38 Prob > F = 0.0000 R-squared = 0.3229 Root MSE = 19.925 (Std. Err. adjusted for 4115 clusters in idcode) ------------------------------------------------------------------------------ | Robust y2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- z1 | .2723216 .0197632 13.78 0.000 .2335751 .3110682 z3 | -.0106124 .0164476 -0.65 0.519 -.0428586 .0216339 x1 | 2.272067 .0998882 22.75 0.000 2.076232 2.467902 x2 | 1.121418 .0749954 14.95 0.000 .9743861 1.268449 _cons | 5.98e-08 2.84e-08 2.11 0.035 4.16e-09 1.15e-07 ------------------------------------------------------------------------------ ( 1) z1 = 0 F( 1, 4114) = 189.87 Prob > F = 0.0000 tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = 0.1258 Total (uncentered) SS = 2344.706897 Uncentered R2 = 0.1258 Residual SS = 2049.727585 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008881 .0018214 -0.49 0.626 -.004458 .0026818 z3 | -.0001671 .0002293 -0.73 0.466 -.0006165 .0002823 x1 | .0406883 .004789 8.50 0.000 .0313021 .0500746 x2 | -.0048113 .0022836 -2.11 0.035 -.0092871 -.0003355 _cons | -6.34e-10 7.71e-10 -0.82 0.411 -2.14e-09 8.77e-10 ------------------------------------------------------------------------------ . timer off 3 . . restore . . *************************************************** . ** Test 2: XTIVREG2, Multiple Varaibles ** . ************************************************** . xtivreg2 y1 (y2 = z1) x1 x2 , fe Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27177 F( 3, 23059) = 1220.64 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = 0.1257 Total (uncentered) SS = 2344.706896 Uncentered R2 = 0.1257 Residual SS = 2049.866562 Root MSE = .2981 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008901 .0008915 -1.00 0.318 -.0026374 .0008571 x1 | .0406984 .0025369 16.04 0.000 .0357263 .0456705 x2 | -.0048145 .0012336 -3.90 0.000 -.0072323 -.0023967 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 236.661 Chi-sq(1) P-val = 0.0000 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 239.084 Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38 15% maximal IV size 8.96 20% maximal IV size 6.66 25% maximal IV size 5.53 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 Included instruments: x1 x2 Excluded instruments: z1 ------------------------------------------------------------------------------ . tsls y1 (y2 = z1) x1 x2 , demean fe(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = 0.1257 Total (uncentered) SS = 2344.706897 Uncentered R2 = 0.1257 Residual SS = 2049.866563 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008901 .0008915 -1.00 0.318 -.0026374 .0008571 x1 | .0406984 .0025369 16.04 0.000 .0357263 .0456705 x2 | -.0048145 .0012336 -3.90 0.000 -.0072323 -.0023967 _cons | -6.35e-10 .0018085 -0.00 1.000 -.0035446 .0035446 ------------------------------------------------------------------------------ . . ******************************* . ** Test 3: Multiple IVs ** . ******************************* . xtivreg2 y1 (y2 = z1 z2) x1 x2 , fe Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27177 F( 3, 23059) = 1220.67 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = 0.1258 Total (uncentered) SS = 2344.706896 Uncentered R2 = 0.1258 Residual SS = 2049.822068 Root MSE = .2981 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008891 .0008914 -1.00 0.319 -.0026363 .000858 x1 | .0406959 .0025368 16.04 0.000 .0357239 .0456679 x2 | -.0048155 .0012335 -3.90 0.000 -.0072332 -.0023978 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 236.669 Chi-sq(2) P-val = 0.0000 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 119.541 Stock-Yogo weak ID test critical values: 10% maximal IV size 19.93 15% maximal IV size 11.59 20% maximal IV size 8.75 25% maximal IV size 7.25 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.042 Chi-sq(1) P-val = 0.8384 ------------------------------------------------------------------------------ Instrumented: y2 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 = z1 z2) x1 x2 , demean fe(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = 0.1258 Total (uncentered) SS = 2344.706897 Uncentered R2 = 0.1258 Residual SS = 2049.822069 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0008891 .0008914 -1.00 0.319 -.0026363 .000858 x1 | .0406959 .0025368 16.04 0.000 .0357239 .0456679 x2 | -.0048155 .0012335 -3.90 0.000 -.0072332 -.0023978 _cons | -6.35e-10 .0018085 -0.00 1.000 -.0035445 .0035445 ------------------------------------------------------------------------------ . . ************************************** . ** Test 4: Multiple Endogenous ** . ************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27177 F( 4, 23058) = 467.69 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035494 -0.11 0.909 -.0073616 .0065519 y3 | -.0500987 .3437882 -0.15 0.884 -.7239112 .6237138 x1 | .0381537 .0178023 2.14 0.032 .003262 .0730455 x2 | -.0045437 .0025412 -1.79 0.074 -.0095243 .000437 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 0.044 Chi-sq(1) P-val = 0.8347 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035494 -0.11 0.909 -.0073616 .0065519 y3 | -.0500987 .3437881 -0.15 0.884 -.7239111 .6237137 x1 | .0381537 .0178023 2.14 0.032 .003262 .0730455 x2 | -.0045437 .0025412 -1.79 0.074 -.0095243 .000437 _cons | -8.73e-10 .0025302 -0.00 1.000 -.0049591 .0049591 ------------------------------------------------------------------------------ . . ************************************** . ** Test 5: Robust ** . ************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe robust Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity Number of obs = 27177 F( 4, 23058) = 409.74 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037642 -0.11 0.914 -.0077826 .0069728 y3 | -.0500987 .3465017 -0.14 0.885 -.7292295 .6290321 x1 | .0381537 .0181649 2.10 0.036 .0025512 .0737563 x2 | -.0045437 .0028309 -1.61 0.108 -.0100922 .0010048 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8362 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) robust tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037642 -0.11 0.914 -.0077826 .0069728 y3 | -.0500987 .3465016 -0.14 0.885 -.7292294 .629032 x1 | .0381537 .0181649 2.10 0.036 .0025512 .0737563 x2 | -.0045437 .0028309 -1.61 0.108 -.0100922 .0010048 _cons | -8.73e-10 .0025302 -0.00 1.000 -.0049591 .0049591 ------------------------------------------------------------------------------ . . ************************************** . ** Test 5: Cluster Robust ** . ************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe robust cluster(idcode) Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 257.20 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433405 -0.15 0.884 -.7230337 .6228363 x1 | .0381537 .0182185 2.09 0.036 .0024461 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8357 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) robust cluster(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433404 -0.15 0.884 -.7230336 .6228362 x1 | .0381537 .0182185 2.09 0.036 .0024462 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 _cons | -8.73e-10 1.81e-09 -0.48 0.629 -4.42e-09 2.67e-09 ------------------------------------------------------------------------------ . . ************************************** . ** Test 5: Small Sample Errors ** . ************************************** . . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe small Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27177 F( 4, 23058) = 467.69 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035497 -0.11 0.909 -.0073626 .0065528 y3 | -.0500987 .343818 -0.15 0.884 -.724005 .6238076 x1 | .0381537 .0178038 2.14 0.032 .0032571 .0730504 x2 | -.0045437 .0025414 -1.79 0.074 -.009525 .0004377 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 0.044 Chi-sq(1) P-val = 0.8347 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) small tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035497 -0.11 0.909 -.0073622 .0065525 y3 | -.0500987 .343818 -0.15 0.884 -.7239695 .6237721 x1 | .0381537 .0178038 2.14 0.032 .0032589 .0730485 x2 | -.0045437 .0025414 -1.79 0.074 -.0095248 .0004374 _cons | -8.73e-10 .0025304 -0.00 1.000 -.0049595 .0049595 ------------------------------------------------------------------------------ . . ******************************************** . ** Test 5: Robust Small Sample Errors ** . ******************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe small robust Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity Number of obs = 27177 F( 4, 23058) = 409.74 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037645 -0.11 0.914 -.0077836 .0069738 y3 | -.0500987 .3465317 -0.14 0.885 -.7293241 .6291267 x1 | .0381537 .0181665 2.10 0.036 .0025462 .0737612 x2 | -.0045437 .0028312 -1.60 0.109 -.0100929 .0010056 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8362 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) small robust tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037645 -0.11 0.914 -.0077832 .0069734 y3 | -.0500987 .3465317 -0.14 0.885 -.7292883 .6290909 x1 | .0381537 .0181665 2.10 0.036 .0025481 .0737594 x2 | -.0045437 .0028311 -1.60 0.109 -.0100926 .0010053 _cons | -8.73e-10 .0025304 -0.00 1.000 -.0049595 .0049595 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Cluster Robust Small Sample Errors ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe small robust cluster(idcode) Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 257.20 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038651 -0.10 0.917 -.0079826 .0071728 y3 | -.0500987 .3434012 -0.15 0.884 -.7233507 .6231533 x1 | .0381537 .0182217 2.09 0.036 .0024293 .0738782 x2 | -.0045437 .0031863 -1.43 0.154 -.0107906 .0017032 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8357 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , demean fe(idcode) small robust cluster(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 2344.706897 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706897 Uncentered R2 = -0.7113 Residual SS = 4012.391281 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038652 -0.10 0.917 -.0079805 .0071707 y3 | -.0500987 .3434074 -0.15 0.884 -.7231649 .6229675 x1 | .0381537 .018222 2.09 0.036 .0024392 .0738683 x2 | -.0045437 .0031864 -1.43 0.154 -.0107889 .0017015 _cons | -8.73e-10 1.81e-09 -0.48 0.629 -4.42e-09 2.67e-09 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg ** . **************************************************** . xtivreg2 y1 (y2 = z1 ) , fe Warning - singleton groups detected. 564 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4121 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27206 F( 1, 23084) = 3.17 Prob > F = 0.0751 Total (centered) SS = 2346.151126 Centered R2 = 0.0514 Total (uncentered) SS = 2346.151126 Uncentered R2 = 0.0514 Residual SS = 2225.547008 Root MSE = .3105 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | .0015398 .000865 1.78 0.075 -.0001556 .0032351 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 186.604 Chi-sq(1) P-val = 0.0000 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 188.116 Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38 15% maximal IV size 8.96 20% maximal IV size 6.66 25% maximal IV size 5.53 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 Excluded instruments: z1 ------------------------------------------------------------------------------ . tsls y1 (y2 = z1 ) , fe(idcode) areg tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27206 Total (centered) SS = 6139.857761 Centered R2 = 0.6375 Total (uncentered) SS = 6139.857761 Uncentered R2 = 0.6375 Residual SS = 2225.547008 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | .0015398 .000865 1.78 0.075 -.0001556 .0032351 _cons | 1.59804 .0470349 33.98 0.000 1.505853 1.690227 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Cluster ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe cluster(idcode) Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 257.20 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433405 -0.15 0.884 -.7230337 .6228363 x1 | .0381537 .0182185 2.09 0.036 .0024461 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8357 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg cluster(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433405 -0.15 0.884 -.7230337 .6228363 x1 | .0381537 .0182185 2.09 0.036 .0024461 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 _cons | 2.096538 3.311457 0.63 0.527 -4.393798 8.586873 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Robust ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe robust Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity Number of obs = 27177 F( 4, 23058) = 409.74 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037642 -0.11 0.914 -.0077826 .0069728 y3 | -.0500987 .3465017 -0.14 0.885 -.7292295 .6290321 x1 | .0381537 .0181649 2.10 0.036 .0025512 .0737563 x2 | -.0045437 .0028309 -1.61 0.108 -.0100922 .0010048 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8362 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg robust tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037642 -0.11 0.914 -.0077826 .0069728 y3 | -.0500987 .3465017 -0.14 0.885 -.7292295 .6290321 x1 | .0381537 .0181649 2.10 0.036 .0025512 .0737563 x2 | -.0045437 .0028309 -1.61 0.108 -.0100922 .0010048 _cons | 2.096538 3.341634 0.63 0.530 -4.452945 8.64602 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Cluster & Robust ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe robust cluster(idcode) Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 257.20 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433405 -0.15 0.884 -.7230337 .6228363 x1 | .0381537 .0182185 2.09 0.036 .0024461 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8357 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg robust cluster(idcode) tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038644 -0.10 0.917 -.007979 .0071693 y3 | -.0500987 .3433405 -0.15 0.884 -.7230337 .6228363 x1 | .0381537 .0182185 2.09 0.036 .0024461 .0738613 x2 | -.0045437 .0031858 -1.43 0.154 -.0107876 .0017003 _cons | 2.096538 3.311457 0.63 0.527 -4.393798 8.586873 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Small SEs ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe small Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics consistent for homoskedasticity only Number of obs = 27177 F( 4, 23058) = 467.69 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035497 -0.11 0.909 -.0073626 .0065528 y3 | -.0500987 .343818 -0.15 0.884 -.724005 .6238076 x1 | .0381537 .0178038 2.14 0.032 .0032571 .0730504 x2 | -.0045437 .0025414 -1.79 0.074 -.009525 .0004377 ------------------------------------------------------------------------------ Underidentification test (Anderson canon. corr. LM statistic): 0.044 Chi-sq(1) P-val = 0.8347 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg small tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0035497 -0.11 0.909 -.0073622 .0065525 y3 | -.0500987 .343818 -0.15 0.884 -.7239697 .6237723 x1 | .0381537 .0178038 2.14 0.032 .0032589 .0730486 x2 | -.0045437 .0025414 -1.79 0.074 -.0095248 .0004374 _cons | 2.096538 3.315631 0.63 0.527 -4.40198 8.595055 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Robust & Small SEs ** . **************************************************** . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe robust small Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity Number of obs = 27177 F( 4, 23058) = 409.74 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037645 -0.11 0.914 -.0077836 .0069738 y3 | -.0500987 .3465317 -0.14 0.885 -.7293241 .6291267 x1 | .0381537 .0181665 2.10 0.036 .0025462 .0737612 x2 | -.0045437 .0028312 -1.60 0.109 -.0100929 .0010056 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8362 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg robust small tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0037645 -0.11 0.914 -.0077832 .0069734 y3 | -.0500987 .3465317 -0.14 0.885 -.7292884 .629091 x1 | .0381537 .0181665 2.10 0.036 .0025481 .0737594 x2 | -.0045437 .0028312 -1.60 0.109 -.0100926 .0010053 _cons | 2.096538 3.341924 0.63 0.530 -4.453513 8.646588 ------------------------------------------------------------------------------ . . **************************************************** . ** Test 5: Areg w/ Cluster & Small SEs ** . **************************************************** . . * The standard errors look off for this. . xtivreg2 y1 (y2 y3 = z1 z2) x1 x2 , fe cluster(idcode) small Warning - singleton groups detected. 569 observation(s) not used. FIXED EFFECTS ESTIMATION ------------------------ Number of groups = 4115 Obs per group: min = 2 avg = 6.6 max = 15 IV (2SLS) estimation -------------------- Estimates efficient for homoskedasticity only Statistics robust to heteroskedasticity and clustering on idcode Number of clusters (idcode) = 4115 Number of obs = 27177 F( 4, 4114) = 257.20 Prob > F = 0.0000 Total (centered) SS = 2344.706896 Centered R2 = -0.7113 Total (uncentered) SS = 2344.706896 Uncentered R2 = -0.7113 Residual SS = 4012.391853 Root MSE = .4171 ------------------------------------------------------------------------------ | Robust y1 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038651 -0.10 0.917 -.0079826 .0071728 y3 | -.0500987 .3434012 -0.15 0.884 -.7233507 .6231533 x1 | .0381537 .0182217 2.09 0.036 .0024293 .0738782 x2 | -.0045437 .0031863 -1.43 0.154 -.0107906 .0017032 ------------------------------------------------------------------------------ Underidentification test (Kleibergen-Paap rk LM statistic): 0.043 Chi-sq(1) P-val = 0.8357 ------------------------------------------------------------------------------ Weak identification test (Cragg-Donald Wald F statistic): 0.022 (Kleibergen-Paap rk Wald F statistic): 0.021 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Source: Stock-Yogo (2005). Reproduced by permission. NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: y2 y3 Included instruments: x1 x2 Excluded instruments: z1 z2 ------------------------------------------------------------------------------ . tsls y1 (y2 y3 = z1 z2) x1 x2 , fe(idcode) areg cluster(idcode) small tsls, Fixed-Effects Two-Stage Instrumental Variable Results Number of clusters () = 4115 Number of obs = 27177 Total (centered) SS = 6133.660264 Centered R2 = 0.3458 Total (uncentered) SS = 6133.660264 Uncentered R2 = 0.3458 Residual SS = 4012.391853 ------------------------------------------------------------------------------ y1 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- y2 | -.0004049 .0038652 -0.10 0.917 -.0079805 .0071707 y3 | -.0500987 .3434075 -0.15 0.884 -.723165 .6229676 x1 | .0381537 .0182221 2.09 0.036 .0024392 .0738683 x2 | -.0045437 .0031864 -1.43 0.154 -.0107889 .0017015 _cons | 2.096538 3.312103 0.63 0.527 -4.395065 8.58814 ------------------------------------------------------------------------------ . . . ********************************************** . * Timing results ** . * timer 1 - time for xtivreg2 ** . * timer 2 - time for tsls ** . * timer 3 - time for tsls with demeaned data** . ********************************************** . . timer list 1: 1.95 / 1 = 1.9480 2: 0.53 / 1 = 0.5290 3: 0.07 / 1 = 0.0660 . . end of do-file . exit,clear