SPREGXT: New Stata Module Econometric Toolkit to Estimate Spatial Panel Regression Models

[Emad Shehata]

Thanks to Professor Kit Baum.

*** TITLE:
SPREGXT: Stata Module Econometric Toolkit to Estimate Spatial Panel Regression Models

*** Author:
Emad Abd Elmessih Shehata
Professor (PhD Economics)
Agricultural Research Center
Agricultural Economics Research Institute
Egypt

Email: [email protected]
WebPage at IDEAS: http://ideas.repec.org/f/psh494.html
WebPage at EconPapers: http://econpapers.repec.org/RAS/psh494.htm

*** SPREGXT Citation:
Shehata, Emad Abd Elmessih (2016)
SPREGXT: "Stata Module Econometric Toolkit to Estimate Spatial Panel Regression Models"

*** Abstract:
spregxt is a Stata Toolkit to estimate Spatial Panel Regression Models: (SAR-SEM-SDM-SAC-GWR-mSTAR-SPGMM-GS2SLS-Tobit) for panel data with be, fe, pa, re Effects, and calculate Panel Autocorrelation, Panel Non Normality, Panel Heteroscedasticity, Panel Error Component Tests, Panel Unit Roots Tests, Identification Restrictions, Hausman Fixed Effects, Specification Error Tests and Model Selection Diagnostic Criteria. Total, Direct, and InDirect Marginal Effects and Elasticities, spregxt can fit continuous and truncated dependent variables models

*** Keywords:
- Regression
- Spatial
- Panel
- Cross Sections-Time Series
- (SAR) MLE Spatial Panel Lag Model
- (SDM) MLE Spatial Panel Durbin Model
- (SEM) MLE Spatial Panel Error Model
- (SAC) MLE Spatial Panel Lag / Error Model "Spatial AutoCorrelation"
- (mSTAR) Multiparametric Spatio Temporal AutoRegressive Regression Model
- (SPGMM) Spatial Panel Autoregressive Generalized Method of Moments Model
- (GS2SLS) Generalized Spatial Panel 2SLS Model
- (GS2SLSAR) Generalized Spatial Panel Autoregressive 2SLS Model
- (OLSXT) Linear Panel Models (Non Spatial)
- (SARXT) Linear Spatial Panel Lag Models (SAR)
- (SDMXT) Linear Spatial Panel Durbin Models (SDM)
- Spatial Between-Effects Panel Regression
- Spatial Fixed-Effects Panel Regression
- Spatial Population Averaged-Effects Panel Regression
- Spatial MLE Random-Effects Panel Regression
- Spatial Amemiya Random-Effects Panel Regression
- Spatial Balestra-Nerlove Random-Effects Panel Regression
- Spatial Fama-MacBeth Panel Regression
- Spatial Hildreth-Houck Random Coefficients Panel Regression
- Spatial Swamy Random Coefficients Panel Regression
- Spatial Geographically Weighted Regressions (GWR)
- Spatial GLS Random-Effects Panel Regression
- Spatial Fuller-Battese GLS Random-Effects Panel Regression
- Spatial Swamy-Arora Random-Effects Panel Regression
- Spatial Trevor Breusch MLE Random-Effects Panel Regression
- Spatial Within-Effects Panel Regression
- Spatial Wallace-Hussain Random-Effects Panel Regression
- Spatial Autocorrelation & Heteroskedasticity GLS Panel Regression
- Spatial Kmenta Homoscedastic GLS AR(1) Panel Regression
- Spatial Parks Full Heteroscedastic Cross-Section GLS AR(1) Panel Regression
- Spatial Heterogeneous Slopes Time Series Panel Regression
- Spatial Corrected Standard Error Panel Regression
- Spatial AR(1) Panel Regression
- Spatial Arellano-Bond Linear Dynamic Panel Regression
- Spatial Han-Philips (2010) Linear Dynamic Panel Regression
- Spatial Arellano-Bond (1991) Linear Dynamic Panel Regression
- Spatial Arellano-Bover/Blundell-Bond (1995
- 1998) System Linear Dynamic Panel
- Spatial Stochastic Frontier Panel Regression
- Spatial Tobit Random-Effects Panel Regression
- Spatial MLE Random-Effects Panel Regression
- Spatial MLE Random-Effects Panel Multiplicative Heteroscedasticity Model
- Aautocorrelation Tests
- Direct Marginal Effects
- Elasticities
- GS2SLS
- GS2SLSAR
- Heteroscedasticity Tests
- Identification Tests
- InDirect Marginal Effects
- Non Normality Tests
- Population-Averaged Effects
- Ramsey RESET
- REgression Specification Error Tests
- Spatial Panel Durbin Model (SDM)
- Spatial Panel Error Model (SEM)
- Spatial Panel Exponential Regression
- Spatial Panel Lag Model (SAR)
- Spatial Panel Multiplicative Heteroscedasticity
- Spatial Panel Weibull Regression
- Total Marginal Effects
- Panel Error Component Tests
- Panel Unit Roots Tests

[Emad Shehata]

*** Example

. clear all
. sysuse spregxt.dta, clear
. spregxt y x1 x2 , nc(7) wmfile(SPWxt) model(sar) mfx(lin) pmfx tests predict(Yh) resid(Ue)

==============================================================================
*** Binary (0/1) Weight Matrix: (49x49) : NC=7 NT=7 (Non Normalized)
------------------------------------------------------------------------------
==============================================================================
* MLE Spatial Panel Lag Normal Model (SAR)
==============================================================================
 y = x1 x2
------------------------------------------------------------------------------
 Sample Size        =          49   |   Cross Sections Number   =           7
 Wald Test          =     54.7117   |   P-Value > Chi2(2)       =      0.0000
 F-Test             =     27.3558   |   P-Value > F(2 , 40)     =      0.0000
 R2  (R-Squared)    =      0.5433   |   Raw Moments R2          =      0.9169
 R2a (Adjusted R2)  =      0.4519   |   Raw Moments R2 Adj      =      0.9003
 Root MSE (Sigma)   =     12.3874   |   Log Likelihood Function =   -187.2901
------------------------------------------------------------------------------
- R2h= 0.5524   R2h Adj= 0.4628  F-Test =   28.38 P-Value > F(2 , 40)  0.0000
- R2r= 0.9169   R2r Adj= 0.9003  F-Test =  169.29 P-Value > F(3 , 40)  0.0000
------------------------------------------------------------------------------
           y |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
y            |
          x1 |  -.2640053   .1025716    -2.57   0.010    -.4650418   -.0629687
          x2 |  -1.599966   .3231151    -4.95   0.000     -2.23326   -.9666723
       _cons |   69.85042   5.446557    12.82   0.000     59.17537    80.52548
-------------+----------------------------------------------------------------
         Rho |  -.0157471    .037723    -0.42   0.676    -.0896829    .0581887
       Sigma |   11.05584    1.11696     9.90   0.000     8.866635    13.24504
------------------------------------------------------------------------------
 LR Test SAR vs. OLS (Rho=0):      0.1743   P-Value > Chi2(1)   0.6764
 Acceptable Range for Rho:        -0.5201   <  Rho  < 0.3115
------------------------------------------------------------------------------

==============================================================================
* Panel Model Selection Diagnostic Criteria - Model= (sar)
==============================================================================
- Log Likelihood Function                   LLF            =   -187.2901
---------------------------------------------------------------------------
- Akaike Information Criterion              (1974) AIC     =    141.5802
- Akaike Information Criterion              (1973) Log AIC =      4.9529
---------------------------------------------------------------------------
- Schwarz Criterion                         (1978) SC      =    158.9663
- Schwarz Criterion                         (1978) Log SC  =      5.0687
---------------------------------------------------------------------------
- Amemiya Prediction Criterion              (1969) FPE     =    162.8422
- Hannan-Quinn Criterion                    (1979) HQ      =    147.9405
- Rice Criterion                            (1984) Rice    =    142.7418
- Shibata Criterion                         (1981) Shibata =    140.6016
- Craven-Wahba Generalized Cross Validation (1979) GCV     =    142.1347
------------------------------------------------------------------------------

==============================================================================
*** Spatial Panel Aautocorrelation Tests - Model= (sar)
*** Binary (0/1) Weight Matrix (W): (Non Normalized)
==============================================================================
  Ho: Error has No Spatial AutoCorrelation
  Ha: Error has    Spatial AutoCorrelation

- GLOBAL Moran MI            =  -0.0792     P-Value > Z(-0.512)   0.6086
- GLOBAL Geary GC            =   1.0330     P-Value > Z(0.235)    0.8140
- GLOBAL Getis-Ords GO       =   0.2262     P-Value > Z(0.512)    0.6086
------------------------------------------------------------------------------
- Moran MI Error Test        =  -0.0886     P-Value > Z(-0.595)   0.9294
------------------------------------------------------------------------------
- LM Error (Burridge)        =   0.3866     P-Value > Chi2(1)     0.5341
- LM Error (Robust)          =   0.1108     P-Value > Chi2(1)     0.7392
------------------------------------------------------------------------------
  Ho: Spatial Lagged Dependent Variable has No Spatial AutoCorrelation
  Ha: Spatial Lagged Dependent Variable has    Spatial AutoCorrelation

- LM Lag (Anselin)           =   0.2758     P-Value > Chi2(1)     0.5995
- LM Lag (Robust)            =   0.0000     P-Value > Chi2(1)     0.9981
------------------------------------------------------------------------------
  Ho: No General Spatial AutoCorrelation
  Ha:    General Spatial AutoCorrelation

- LM SAC (LMErr+LMLag_R)     =   0.3866     P-Value > Chi2(2)     0.8242
- LM SAC (LMLag+LMErr_R)     =   0.3866     P-Value > Chi2(2)     0.8242
------------------------------------------------------------------------------

==============================================================================
*** Panel Unit Roots Tests - Model= (sar)
==============================================================================
  Ho: All Panels are Stationary - Ha: Some Panels Have Unit Roots

- Hadri Z Test (No Trend - No Robust) =   1.9540   P-Value > Z(0,1)   0.0254
- Hadri Z Test (No Trend -    Robust) =   2.0098   P-Value > Z(0,1)   0.0222
- Hadri Z Test (   Trend - No Robust) =   0.9539   P-Value > Z(0,1)   0.1701
- Hadri Z Test (   Trend -    Robust) =   0.6572   P-Value > Z(0,1)   0.2555
------------------------------------------------------------------------------
==============================================================================
* (1)  (DF):           Dickey-Fuller   Test
* (2) (ADF): Augmented Dickey-Fuller   Test
* (3) (APP): Augmented Phillips-Perron Test
--------------------------------------------------
  Ho: All Panels Have Unit Roots  (Non stationary)
  Ha: At Least One Panel is Stationary
------------------------------------------------------------------------------
  Ho: Non Stationary [0.05, 0.01 < P-Value]
  Ha:     Stationary [0.05, 0.01 > P-Value]
------------------------------------------------------------------------------
*** (1) Dickey-Fuller (DF) Test:
--------------------------------------------------
-  DF Test: [Lag = 0] (No Trend)  =  -2.1511     P-Value > Z(0,1)    0.0157
* Since [.05 > 0.0157]:   Variable (y) has Stationary Process
------------------------------------------------------------------------------
-  DF Test: [Lag = 0] (   Trend)  =  -1.7131     P-Value > Z(0,1)    0.0433
* Since [.05 > 0.0433]:   Variable (y) has Stationary Process
------------------------------------------------------------------------------

------------------------------------------------------------------------------
*** (2) Augmented Dickey-Fuller (ADF) Test:
--------------------------------------------------
- ADF Test: [Lag = 1] (No Trend)  =   0.6539     P-Value > Z(0,1)    0.7434
* Since [.05 < 0.7434]:   Variable (y) has Non Stationary (Unit Roots)
------------------------------------------------------------------------------
- ADF Test: [Lag = 1] (   Trend)  =  -3.8085     P-Value > Z(0,1)    0.0001
* Since [.05 > 0.0001]:   Variable (y) has Stationary Process
------------------------------------------------------------------------------

------------------------------------------------------------------------------
*** (3) Augmented Phillips-Perron (APP) Test:
--------------------------------------------------
- APP Test: [Lag = 1] (No Trend)  =  -2.8930     P-Value > Z(0,1)    0.0019
* Since [.05 > 0.0019]:   Variable (y) has Stationary Process
------------------------------------------------------------------------------
- APP Test: [Lag = 1] (   Trend)  =  -4.5522     P-Value > Z(0,1)    0.0000
* Since [.05 > 0.0000]:   Variable (y) has Stationary Process
------------------------------------------------------------------------------

==============================================================================
*** Panel Error Component Tests - Model= (sar)
==============================================================================
* Panel Random Effects Tests
  Ho: No AR(1) Autocorrelation - Ha: AR(1) Autocorrelation
  Ho: Pooled OLS    - No Significance Difference among Panels
  Ha: Random Effect -    Significance Difference among Panels

- Breusch-Pagan  LM Test -Two Side      =  20.5375  P-Value > Chi2(1)   0.0000
- Breusch-Pagan ALM Test -Two Side      =  10.7351  P-Value > Chi2(1)   0.0011
------------------------------------------------------------------------------
- Sosa-Escudero-Yoon  LM Test -One Side =   4.5318  P-Value > Chi2(1)   0.0333
- Sosa-Escudero-Yoon ALM Test -One Side =   3.2764  P-Value > Chi2(1)   0.0703
------------------------------------------------------------------------------
- Baltagi-Li  LM Autocorrelation Test   =  10.8752  P-Value > Chi2(1)   0.0010
- Baltagi-Li ALM Autocorrelation Test   =   1.0729  P-Value > Chi2(1)   0.3003
------------------------------------------------------------------------------
- Baltagi-Li LM AR(1) Joint Test        =  21.6103  P-Value > Chi2(2)   0.0000
------------------------------------------------------------------------------

* Contemporaneous Correlations Across Cross Sctions Test
  Ho: No Contemporaneous Correlations (Independence) - (Pooled OLS)
  Ha:    Contemporaneous Correlations (Dependence)   - (Panel)

- Breusch-Pagan Diagonal Covariance Matrix LM Test=  32.3077 P>Chi2(21) 0.0545
- Breusch-Pagan Cross-Section Independence LM Test=  26.4443 P>Chi2(21) 0.1900
------------------------------------------------------------------------------
    LM= Lagrange Multiplier ; ALM = Adjusted Lagrange Multiplier
------------------------------------------------------------------------------

==============================================================================
*** Panel Serial Autocorrelation Tests - Model= (sar)
==============================================================================
  Ho: No AR(1) Panel AutoCorrelation - Ha: AR(1) Panel AutoCorrelation

- Durbin  h Test (Lag DepVar)             =   4.1863  P-Value > Z(0,1)  0.0000
- Harvey LM Test (Lag DepVar)             =  17.5248  P-Value > Chi2(1) 0.0000
------------------------------------------------------------------------------
- Panel Rho Value                         =   0.4360
- Durbin-Watson Test                      =   1.0212  df: (3 , 49)
- Von Neumann Ratio Test                  =   1.0424  df: (3 , 49)
- Box-Pierce LM Test                      =   9.3135  P-Value > Chi2(1) 0.0023
- Z Test                                  =   3.0518  P-Value > Z(0,1)  0.0023
------------------------------------------------------------------------------
- Durbin m Test (drop 1 cs obs)           =  10.3289  P-Value > F(1,38) 0.0027
- Durbin m Test (keep 1 cs obs)           =  10.9844  P-Value > F(1,45) 0.0018
------------------------------------------------------------------------------
- Breusch-Godfrey LM Test (drop 1 cs obs) =   9.0507  P-Value > Chi2(1) 0.0026
- Breusch-Godfrey LM Test (keep 1 cs obs) =   9.6238  P-Value > Chi2(1) 0.0019
------------------------------------------------------------------------------
- Breusch-Pagan-Godfrey Z (keep 1 nt obs) =   2.8841  P-Value > Z(0,1)  0.0039
- Breusch-Pagan-Godfrey Z (drop 1 cs obs) =   3.0084  P-Value > Z(0,1)  0.0026
- Breusch-Pagan-Godfrey Z (keep 1 cs obs) =   3.1022  P-Value > Z(0,1)  0.0019
------------------------------------------------------------------------------
- Baltagi LM Test                         =  10.8658  P-Value > Chi2(1) 0.0010
- Baltagi  Z Test                         =   3.2963  P-Value > Z(0,1)  0.0010
------------------------------------------------------------------------------
- Wooldridge  F Test                      =   8.6440  P-Value > F(1, 6) 0.0259
- Wooldridge LM Test                      =   7.0879  P-Value > Chi2(1) 0.0078
------------------------------------------------------------------------------

==============================================================================
*** Panel Heteroscedasticity Tests - Model= (sar)
==============================================================================
  Ho: Panel Homoscedasticity - Ha: Panel Heteroscedasticity

- Engle LM ARCH Test AR(1): E2 = E2_1   =   0.4949   P-Value > Chi2(1)  0.4818
------------------------------------------------------------------------------
- Hall-Pagan LM Test:   E2 = Yh         =   0.7118   P-Value > Chi2(1)  0.3988
- Hall-Pagan LM Test:   E2 = Yh2        =   0.3832   P-Value > Chi2(1)  0.5359
- Hall-Pagan LM Test:   E2 = LYh2       =   0.8228   P-Value > Chi2(1)  0.3644
------------------------------------------------------------------------------
- Harvey LM Test:    LogE2 = X          =   4.7496   P-Value > Chi2(2)  0.0930
- Wald Test:         LogE2 = X          =  11.7191   P-Value > Chi2(1)  0.0006
- Glejser LM Test:     |E| = X          =   5.6568   P-Value > Chi2(2)  0.0591
- Breusch-Godfrey Test:  E = E_1 X      =   7.5457   P-Value > Chi2(1)  0.0060
------------------------------------------------------------------------------
- Machado-Santos-Silva Test: Ev=Yh Yh2  =   2.7231   P-Value > Chi2(2)  0.2563
- Machado-Santos-Silva Test: Ev=X       =   6.0504   P-Value > Chi2(2)  0.0485
------------------------------------------------------------------------------
- White Test - Koenker(R2): E2 = X      =   6.6105   P-Value > Chi2(2)  0.0367
- White Test - B-P-G (SSR): E2 = X      =   9.8628   P-Value > Chi2(2)  0.0072
------------------------------------------------------------------------------
- White Test - Koenker(R2): E2 = X X2   =   7.3435   P-Value > Chi2(4)  0.1188
- White Test - B-P-G (SSR): E2 = X X2   =  10.9566   P-Value > Chi2(4)  0.0271
------------------------------------------------------------------------------
- White Test - Koenker(R2): E2 = X X2 XX=  20.8630   P-Value > Chi2(5)  0.0009
- White Test - B-P-G (SSR): E2 = X X2 XX=  31.1276   P-Value > Chi2(5)  0.0000
------------------------------------------------------------------------------
- Cook-Weisberg LM Test: E2/S2n = Yh    =   1.0620   P-Value > Chi2(1)  0.3028
- Cook-Weisberg LM Test: E2/S2n = X     =   9.8628   P-Value > Chi2(2)  0.0072
------------------------------------------------------------------------------
*** Single Variable Tests: ***
* Cook-Weisberg LM Test: E2/Sig2
- x1                                   =   1.2426    P-Value > Chi2(1)  0.2650
- x2                                   =   3.9428    P-Value > Chi2(1)  0.0471
------------------------------------------------------------------------------
*** Single Variable Tests: ***
* King LM Test:
- x1                                   =   0.2585    P-Value > Chi2(1)  0.6112
- x2                                   =   3.9478    P-Value > Chi2(1)  0.0469
------------------------------------------------------------------------------

==============================================================================
* Panel Groupwise Heteroscedasticity Tests
==============================================================================
  Ho: Panel Homoscedasticity - Ha: Panel Groupwise Heteroscedasticity

- Lagrange Multiplier LM Test   =      7.3373    P-Value > Chi2(6)   0.2908
- Likelihood Ratio LR Test      =      7.1253    P-Value > Chi2(6)   0.3094
- Wald Test                     =     12.4812    P-Value > Chi2(7)   0.0858
------------------------------------------------------------------------------

==============================================================================
* Panel Non Normality Tests - Model= (sar)
==============================================================================
 Ho: Normality - Ha: Non Normality
------------------------------------------------------------------------------
*** Non Normality Tests:
- Jarque-Bera LM Test                  =   1.6062     P-Value > Chi2(2) 0.4479
- White IM Test                        =   5.8155     P-Value > Chi2(2) 0.0546
- Doornik-Hansen LM Test               =   3.5054     P-Value > Chi2(2) 0.1733
- Geary LM Test                        =  -2.7412     P-Value > Chi2(2) 0.2540
- Anderson-Darling Z Test              =   0.3237     P > Z( 0.118)     0.5472
- D'Agostino-Pearson LM Test           =   2.5494     P-Value > Chi2(2) 0.2795
------------------------------------------------------------------------------
*** Skewness Tests:
- Srivastava LM Skewness Test          =   0.6074     P-Value > Chi2(1) 0.4358
- Small LM Skewness Test               =   0.7396     P-Value > Chi2(1) 0.3898
- Skewness Z Test                      =  -0.8600     P-Value > Chi2(1) 0.3898
------------------------------------------------------------------------------
*** Kurtosis Tests:
- Srivastava  Z Kurtosis Test          =   0.9994     P-Value > Z(0,1)  0.3176
- Small LM Kurtosis Test               =   1.8098     P-Value > Chi2(1) 0.1785
- Kurtosis Z Test                      =   1.3453     P-Value > Chi2(1) 0.1785
------------------------------------------------------------------------------
    Skewness Coefficient = -0.2727     - Standard Deviation =  0.3398
    Kurtosis Coefficient =  3.6994     - Standard Deviation =  0.6681
------------------------------------------------------------------------------
    Runs Test: (16) Runs -  (24) Positives - (25) Negatives
    Standard Deviation Runs Sig(k) =  3.4619 , Mean Runs E(k) = 25.4898
    95% Conf. Interval [E(k)+/- 1.96* Sig(k)] = (18.7045 , 32.2751 )
------------------------------------------------------------------------------

==============================================================================
***  REgression Specification Error Tests (RESET) - (Model= sar)
==============================================================================
 Ho: Model is Specified  -  Ha: Model is Misspecified
------------------------------------------------------------------------------
* Ramsey Specification ResetF Test
- Ramsey RESETF1 Test: Y= X Yh2         =   2.723  P-Value > F(1,  45) 0.1059
- Ramsey RESETF2 Test: Y= X Yh2 Yh3     =   1.344  P-Value > F(2,  44) 0.2713
- Ramsey RESETF3 Test: Y= X Yh2 Yh3 Yh4 =   1.922  P-Value > F(3,  43) 0.1402
------------------------------------------------------------------------------
* DeBenedictis-Giles Specification ResetL Test
- Debenedictis-Giles ResetL1 Test       =   1.585  P-Value > F(2, 44)  0.2164
- Debenedictis-Giles ResetL2 Test       =   1.244  P-Value > F(4, 42)  0.3073
- Debenedictis-Giles ResetL3 Test       =   1.273  P-Value > F(6, 40)  0.2915
------------------------------------------------------------------------------
* DeBenedictis-Giles Specification ResetS Test
- Debenedictis-Giles ResetS1 Test       =   1.754  P-Value > F(2, 44)  0.1850
- Debenedictis-Giles ResetS2 Test       =   2.189  P-Value > F(4, 42)  0.0867
- Debenedictis-Giles ResetS3 Test       =   1.416  P-Value > F(6, 40)  0.2323
------------------------------------------------------------------------------

==============================================================================
*** Multicollinearity Diagnostic Tests - Model= (sar)
==============================================================================

* Correlation Matrix
+------------------------------------------+
|   Variable |       y |      x1 |      x2 |
|------------+---------+---------+---------|
|          y |   1.000 |         |         |
|         x1 |  -0.574 |   1.000 |         |
|         x2 |  -0.696 |   0.500 |   1.000 |
+------------------------------------------+

             |        y       x1       x2
-------------+---------------------------
           y |   1.0000
             |
             |
          x1 |  -0.5745*  1.0000
             |   0.0000
             |
          x2 |  -0.6956*  0.4999*  1.0000
             |   0.0000   0.0003
             |

* Multicollinearity Diagnostic Criteria
+----------------------------------------------------------------------------+
| Variable | Eigenval | C_Number |  C_Index |      VIF |    1/VIF |  R2_xi,X |
|----------+----------+----------+----------+----------+----------+----------|
|       x1 |   1.4999 |   1.0000 |   1.0000 |   1.3331 |   0.7501 |   0.2499 |
|       x2 |   0.5001 |   2.9990 |   1.7318 |   1.3331 |   0.7501 |   0.2499 |
+----------------------------------------------------------------------------+

* Farrar-Glauber Multicollinearity Tests
  Ho: No Multicollinearity - Ha: Multicollinearity
--------------------------------------------------

* (1) Farrar-Glauber Multicollinearity Chi2-Test:
    Chi2 Test =   13.3697    P-Value > Chi2(1) 0.0003

* (2) Farrar-Glauber Multicollinearity F-Test:
+------------------------------------------------------------------------+
|   Variable |       F_Test |          DF1 |          DF2 |      P_Value |
|------------+--------------+--------------+--------------+--------------|
|         x1 |       15.657 |       47.000 |        2.000 |        0.062 |
|         x2 |       15.657 |       47.000 |        2.000 |        0.062 |
+------------------------------------------------------------------------+

* (3) Farrar-Glauber Multicollinearity t-Test:
+------------------------------------+
|   Variable |        x1 |        x2 |
|------------+-----------+-----------|
|         x1 |         . |           |
|------------+-----------+-----------|
|         x2 |     3.957 |         . |
+------------------------------------+

* |X'X| Determinant:
  |X'X| = 0 Multicollinearity - |X'X| = 1 No Multicollinearity
  |X'X| Determinant:       (0 < 0.7501 < 1)
---------------------------------------------------------------

* Theil R2 Multicollinearity Effect:
  R2 = 0 No Multicollinearity - R2 = 1 Multicollinearity
     - Theil R2:           (0 < 0.2706 < 1)
---------------------------------------------------------------

* Multicollinearity Range:
  Q = 0 No Multicollinearity - Q = 1 Multicollinearity
     - Gleason-Staelin Q0: (0 < 0.4999 < 1)
    1- Heo Range Q1:       (0 < 0.1974 < 1)
    2- Heo Range Q2:       (0 < 0.2499 < 1)
    3- Heo Range Q3:       (0 < 0.1339 < 1)
    4- Heo Range Q4:       (0 < 0.6494 < 1)
    5- Heo Range Q5:       (0 < 0.2372 < 1)
    6- Heo Range Q6:       (0 < 0.2499 < 1)
------------------------------------------------------------------------------

==============================================================================
*** Linear vs Log-Linear Functional Form Tests - (Model= OLS)
==============================================================================
 (1) R-squared
      Linear  R2                   =    0.5524
      Log-Log R2                   =    0.3749
---------------------------------------------------------------------------
 (2) Log Likelihood Function (LLF)
      LLF - Linear                 = -187.3772
      LLF - Log-Log                = -224.3892
---------------------------------------------------------------------------
 (3) Antilog R2
      Linear  vs Log-Log: R2Lin    =    0.5269
      Log-Log vs Linear : R2log    =    0.1208
---------------------------------------------------------------------------
 (4) Box-Cox Test                  =   36.2567   P-Value > Chi2(1)   0.0000
      Ho: Choose Log-Log Model - Ha: Choose Linear  Model
---------------------------------------------------------------------------
 (5) Bera-McAleer BM Test
      Ho: Choose Linear  Model     =    2.3994   P-Value > F(1, 44)  0.1285
      Ho: Choose Log-Log Model     =    0.8078   P-Value > F(1, 45)  0.3736
---------------------------------------------------------------------------
 (6) Davidson-Mackinnon PE Test
      Ho: Choose Linear  Model     =    2.3994   P-Value > F(1, 44)  0.1285
      Ho: Choose Log-Log Model     =    0.8078   P-Value > F(1, 45)  0.3736
------------------------------------------------------------------------------

==============================================================================

* Beta, Total, Direct, and InDirect (Model= sar): Linear: Marginal Effect *

+-------------------------------------------------------------------------------+
|     Variable |    Beta(B) |      Total |     Direct |   InDirect |       Mean |
|--------------+------------+------------+------------+------------+------------|
|y             |            |            |            |            |            |
|           x1 |    -0.2640 |    -0.2638 |    -0.2758 |     0.0120 |    38.4362 |
|           x2 |    -1.6000 |    -1.5989 |    -1.6716 |     0.0727 |    14.3749 |
+-------------------------------------------------------------------------------+

* Beta, Total, Direct, and InDirect (Model= sar): Linear: Elasticity *

+-------------------------------------------------------------------------------+
|     Variable |   Beta(Es) |      Total |     Direct |   InDirect |       Mean |
|--------------+------------+------------+------------+------------+------------|
|           x1 |    -0.2889 |    -0.2887 |    -0.3018 |     0.0131 |    38.4362 |
|           x2 |    -0.6547 |    -0.6543 |    -0.6840 |     0.0298 |    14.3749 |
+-------------------------------------------------------------------------------+
 Mean of Dependent Variable =     35.1288

------------------------------------------------------------------------------
*** P-Value: Total, Direct, and InDirect Marginal Effect ***
------------------------------------------------------------------------------

*** (1) Total Marginal Effect ***
------------------------------------------------------------------------------
       Total |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
y            |
          x1 |  -.2638216   .1025716    -2.57   0.014    -.4711265   -.0565168
          x2 |  -1.598853   .3231151    -4.95   0.000    -2.251893   -.9458132
------------------------------------------------------------------------------

*** (2) Direct Marginal Effect ***
------------------------------------------------------------------------------
      Direct |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
y            |
          x1 |  -.2758194   .1025716    -2.69   0.010    -.4831243   -.0685146
          x2 |  -1.671564   .3231151    -5.17   0.000    -2.324604   -1.018524
------------------------------------------------------------------------------

*** (3) InDirect Marginal Effect ***
------------------------------------------------------------------------------
    InDirect |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
y            |
          x1 |   .0119978   .1025716     0.12   0.907    -.1953071    .2193026
          x2 |   .0727109   .3231151     0.23   0.823    -.5803291    .7257508
------------------------------------------------------------------------------

.

[Steven Archambault]

Thank you for this! What can be done with an unbalanced panel? I have a series of well measurements over many years, however these measurements don't always occur annually. I would like to analyze these spatially, but imputing dependent variable (well depth) seems wrong. Any thoughts would be appreciated! Thank you!

[Biswa Swarup]

Dear Prof. Emad Shehata,

I am referring to the documentation of SPREGXT. Can you please advise the paper or Book on Spatial Econometrics which will help to relate the various models SAR, SEM, SDM, General Spatial Models discussed in the help file.

Also, I need a few clarifications.


1. Specifically, what is the distinction between Linear SDM model and MLE SDM model. Is it that the Linear modes is estimated through OLS and gives biased coefficient estimates and as such, estimation of SDM through MLE is desirable.

2. The help manual states that direct and indirect marginal effects can be computed only for sar, sdm, sac models. Is it possible to get the direct and indirect effects from the estimation of general spatial model.

3. Is there any way to estimate the SDEM model for estimating local spillovers.

Thanking you.

Regards,

Biswa

---

---