I have different result in using rhausman vs xtoverid in Hausman Test for Data Panel. Could you help me to solve with the best model?
. xtset ID YEAR
Panel variable: ID (strongly balanced)
Time variable: YEAR, 2012 to 2021
Delta: 1 unit
Is the best model Fixed Effect or Random Effect?
. xtset ID YEAR
Panel variable: ID (strongly balanced)
Time variable: YEAR, 2012 to 2021
Delta: 1 unit
| . xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, re |
| Random-effects GLS regression Number of obs = 820 |
| Group variable: ID Number of groups = 82 |
| R-squared: Obs per group: |
| Within = 0.0582 min = 10 |
| Between = 0.1582 avg = 10.0 |
| Overall = 0.1327 max = 10 |
| Wald chi2(7) = 60.24 |
| corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 |
| TOBINSQ Coefficient Std. err. z P>z [95% conf. interval] |
| MVAIC .1984099 .0394648 5.03 0.000 .1210604 .2757594 |
| DPR -.0483234 .0915305 -0.53 0.598 -.2277199 .1310731 |
| DER -.7312765 .1455461 -5.02 0.000 -1.016542 -.4460115 |
| SIZE .1176504 .0785496 1.50 0.134 -.036304 .2716048 |
| AGE .1241954 .2387606 0.52 0.603 -.3437668 .5921577 |
| GDP -.2313633 1.105906 -0.21 0.834 -2.3989 1.936174 |
| INFL 5.863882 2.525215 2.32 0.020 .9145527 10.81321 |
| _cons -.2438746 .8669913 -0.28 0.778 -1.943146 1.455397 |
| sigma_u 1.1853203 |
| sigma_e .79635106 |
| rho .68900172 (fraction of variance due to u_i) |
| . xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, fe |
| Fixed-effects (within) regression Number of obs = 820 |
| Group variable: ID Number of groups = 82 |
| R-squared: Obs per group: |
| Within = 0.0609 min = 10 |
| Between = 0.0993 avg = 10.0 |
| Overall = 0.0904 max = 10 |
| F(7,731) = 6.77 |
| corr(u_i, Xb) = 0.0566 Prob > F = 0.0000 |
| TOBINSQ Coefficient Std. err. t P>t [95% conf. interval] |
| MVAIC .1470755 .0411792 3.57 0.000 .0662318 .2279191 |
| DPR -.1254703 .0917153 -1.37 0.172 -.3055271 .0545865 |
| DER -.7791355 .156089 -4.99 0.000 -1.085572 -.4726994 |
| SIZE .1676709 .1490193 1.13 0.261 -.1248859 .4602277 |
| AGE -.0902883 .3840589 -0.24 0.814 -.8442783 .6637016 |
| GDP -.3076789 1.111873 -0.28 0.782 -2.490524 1.875166 |
| INFL 5.527369 2.568913 2.15 0.032 .4840413 10.5707 |
| _cons .6033705 1.25904 0.48 0.632 -1.868395 3.075136 |
| sigma_u 1.4100819 |
| sigma_e .79635106 |
| rho .75817975 (fraction of variance due to u_i) |
| F test that all u_i=0: F(81, 731) = 25.58 Prob > F = 0.0000 |
| . reg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL |
| Source SS df MS Number of obs = 820 |
| F(7, 812) = 32.45 |
| Model 497.257251 7 71.0367501 Prob > F = 0.0000 |
| Residual 1777.33082 812 2.18883105 R-squared = 0.2186 |
| Adj R-squared = 0.2119 |
| Total 2274.58807 819 2.77727481 Root MSE = 1.4795 |
| TOBINSQ Coefficient Std. err. t P>t [95% conf. interval] |
| MVAIC .467226 .0421503 11.08 0.000 .3844896 .5499624 |
| DPR .8405944 .13371 6.29 0.000 .5781364 1.103052 |
| DER -.6387247 .1347102 -4.74 0.000 -.903146 -.3743034 |
| SIZE .0544056 .0371676 1.46 0.144 -.0185504 .1273616 |
| AGE .2278863 .1176803 1.94 0.053 -.0031072 .4588798 |
| GDP -.5932197 1.918164 -0.31 0.757 -4.358363 3.171924 |
| INFL 7.58951 3.723769 2.04 0.042 .2801622 14.89886 |
| _cons -2.305617 .463206 -4.98 0.000 -3.21484 -1.396395 |
| . estimates store fe |
| . |
| . estimates store re |
| . |
| . estimate store ols |
| . |
| . estimates table fe re ols, star stats( n r2 r2_a) |
| Variable fe re ols |
| MVAIC .46722603*** .46722603*** .46722603*** |
| DPR .84059438*** .84059438*** .84059438*** |
| DER -.63872466*** -.63872466*** -.63872466*** |
| SIZE .05440562 .05440562 .05440562 |
| AGE .22788628 .22788628 .22788628 |
| GDP -.59321973 -.59321973 -.59321973 |
| INFL 7.5895101* 7.5895101* 7.5895101* |
| _cons -2.3056174*** -2.3056174*** -2.3056174*** |
| n |
| r2 .2186142 .2186142 .2186142 |
| r2_a .21187812 .21187812 .21187812 |
| Legend: * p<0.05;> |
| . hausman fe re |
| Note: the rank of the differenced variance matrix (0) does not equal the number of coefficients being tested (7); be sure this is what you |
| expect, or there may be problems computing the test. Examine the output of your estimators for anything unexpected and possibly |
| consider scaling your variables so that the coefficients are on a similar scale. |
| Coefficients ---- |
| (b) (B) (b-B) sqrt(diag(V_b-V_B)) |
| fe re Difference Std. err. |
| MVAIC .467226 .467226 0 0 |
| DPR .8405944 .8405944 0 0 |
| DER -.6387247 -.6387247 0 0 |
| SIZE .0544056 .0544056 0 0 |
| AGE .2278863 .2278863 0 0 |
| GDP -.5932197 -.5932197 0 0 |
| INFL 7.58951 7.58951 0 0 |
| b = Consistent under H0 and Ha; obtained from regress. |
| B = Inconsistent under Ha, efficient under H0; obtained from regress. |
| Test of H0: Difference in coefficients not systematic |
| chi2(0) = (b-B)'[(V_b-V_B)^(-1)](b-B) |
| = 0.00 |
| Prob > chi2 = . |
| (V_b-V_B is not positive definite) |
| . quietly xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, re |
| . |
| . estimates store re |
| . |
| . quietly xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, fe |
| . |
| . estimates store fe |
| . |
| . |
| . |
| . hausman fe re |
| Coefficients ---- |
| (b) (B) (b-B) sqrt(diag(V_b-V_B)) |
| fe re Difference Std. err. |
| MVAIC .1470755 .1984099 -.0513344 .0117584 |
| DPR -.1254703 -.0483234 -.0771469 .0058189 |
| DER -.7791355 -.7312765 -.047859 .0563925 |
| SIZE .1676709 .1176504 .0500205 .1266361 |
| AGE -.0902883 .1241954 -.2144838 .3008232 |
| GDP -.3076789 -.2313633 -.0763156 .1150308 |
| INFL 5.527369 5.863882 -.3365128 .4718132 |
| b = Consistent under H0 and Ha; obtained from xtreg. |
| B = Inconsistent under Ha, efficient under H0; obtained from xtreg. |
| Test of H0: Difference in coefficients not systematic |
| chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) |
| = -1261.06 |
| Warning: chi2 < 0 ==> model fitted on these data |
| fails to meet the asymptotic assumptions |
| of the Hausman test; see suest for a |
| generalized test. |
| . *Cluster-Robust Hausman Test |
| . rhausman fe re, cluster |
| bootstrap in progress |
| 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 |
| .................................................. 50 |
| .................................................. 100 |
| Cluster-Robust Hausman Test |
| (based on 100 bootstrap repetitions) |
| b1: obtained from xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, fe |
| b2: obtained from xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, re |
| Test: Ho: difference in coefficients not systematic |
| chi2(7) = (b1-b2)' * [V_bootstrapped(b1-b2)]^(-1) * (b1-b2) |
| = 3.44 |
| Prob>chi2 = 0.8414 |
| . xtreg TOBINSQ MVAIC DPR DER SIZE AGE GDP INFL, re |
| Random-effects GLS regression Number of obs = 820 |
| Group variable: ID Number of groups = 82 |
| R-squared: Obs per group: |
| Within = 0.0582 min = 10 |
| Between = 0.1582 avg = 10.0 |
| Overall = 0.1327 max = 10 |
| Wald chi2(7) = 60.24 |
| corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 |
| TOBINSQ Coefficient Std. err. z P>z [95% conf. interval] |
| MVAIC .1984099 .0394648 5.03 0.000 .1210604 .2757594 |
| DPR -.0483234 .0915305 -0.53 0.598 -.2277199 .1310731 |
| DER -.7312765 .1455461 -5.02 0.000 -1.016542 -.4460115 |
| SIZE .1176504 .0785496 1.50 0.134 -.036304 .2716048 |
| AGE .1241954 .2387606 0.52 0.603 -.3437668 .5921577 |
| GDP -.2313633 1.105906 -0.21 0.834 -2.3989 1.936174 |
| INFL 5.863882 2.525215 2.32 0.020 .9145527 10.81321 |
| _cons -.2438746 .8669913 -0.28 0.778 -1.943146 1.455397 |
| sigma_u 1.1853203 |
| sigma_e .79635106 |
| rho .68900172 (fraction of variance due to u_i) |
| . xtoverid |
| Test of overidentifying restrictions: fixed vs random effects |
| Cross-section time-series model: xtreg re |
| Sargan-Hansen statistic 32.538 Chi-sq(7) P-value = 0.0000 |
Is the best model Fixed Effect or Random Effect?
Comment