A) Nominal rate of assistance: 1960 -- 2000 --------------------- Senstivity specification (1a) ---------------------------- ------- Baseline output growth regression (specification 5a in distortthr09.out) plus NRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 34.30766 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.29590 LM-test for no threshold 23.53855 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.070962354 0.035046912 LS 0.016327274 0.0033139267 LPRIMARY 0.0098718171 0.0046757421 LPOP -0.055396187 0.016512993 LRGDP0 -0.012291494 0.0029836517 NRA 0.0033737411 0.0028852225 Observations 211.00000 Degrees of Freedom 205.00000 Sum of Squared Errors 0.08068 Residual Variance 0.00039 R-squared 0.26528 Heteroskedasticity Test (P-Value) 0.00052 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.24237 95% Confidence Interval: [ 6.450786 , 9.133988 ] Sum of Squared Errors 0.06654 Residual Variance 0.00033 Joint R-Squared 0.39402 Heteroskedasticity Test (p-value) 0.00894 _____________________________________________ Regime 1: LRGDP0 <= 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.10546288 0.13850789 LS 0.012931301 0.0043524280 LPRIMARY 0.0013770613 0.0067767354 LPOP -0.093301443 0.045165661 LRGDP0 -0.021466275 0.013541270 NRA 0.031497789 0.013132690 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.38495756 0.17216042 LS 0.0042502084 0.022114123 LPRIMARY -0.011905340 0.015005889 LPOP -0.18270075 0.0012847837 LRGDP0 -0.048154111 0.0071258862 NRA 0.0056023457 0.059338552 Observations 61.00000 Degrees of Freedom 55.00000 Sum of Squared Errors 0.02356 Residual Variance 0.00043 R-squared 0.24213 _____________________________________________ Regime 2: LRGDP0 > 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.0022482697 0.034725436 LS 0.019687981 0.0037905731 LPRIMARY 0.023107437 0.0076292877 LPOP -0.053784737 0.016395936 LRGDP0 -0.020384322 0.0029733471 NRA 0.0036721673 0.0025559503 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.11671622 0.24968186 LS -0.0098758989 0.028442096 LPRIMARY -0.052653727 0.069131149 LPOP -0.087297342 0.017654649 LRGDP0 -0.026905680 -0.0048392979 NRA -0.0019510801 0.0088681756 Observations 150.00000 Degrees of Freedom 144.00000 Sum of Squared Errors 0.04299 Residual Variance 0.00030 R-squared 0.35515 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.070962354 0.035046912 LS 0.016327274 0.0033139267 LPRIMARY 0.0098718171 0.0046757421 LPOP -0.055396187 0.016512993 LRGDP0 -0.012291494 0.0029836517 NRA 0.0033737411 0.0028852225 Observations 211.00000 Degrees of Freedom 205.00000 Sum of Squared Errors 0.08068 Residual Variance 0.00039 R-squared 0.26528 Heteroskedasticity Test (P-Value) 0.00052 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.29590 95% Confidence Interval: [ 0.067800 , 0.698300 ] Sum of Squared Errors 0.06668 Residual Variance 0.00034 Joint R-Squared 0.39276 Heteroskedasticity Test (p-value) 0.00011 _____________________________________________ Regime 1: NRA <= 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.11419192 0.045986870 LS 0.015126893 0.0031969854 LPRIMARY 0.0071590382 0.0044523921 LPOP -0.072997565 0.018164851 LRGDP0 -0.012152525 0.0027449627 NRA 0.027294427 0.010819289 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.22961857 0.0069566188 LS 0.0085895078 0.021709579 LPRIMARY -0.0028137165 0.016821878 LPOP -0.11732023 -0.024249080 LRGDP0 -0.017800399 -0.0045812084 NRA 0.0024887130 0.056802748 Observations 156.00000 Degrees of Freedom 150.00000 Sum of Squared Errors 0.06103 Residual Variance 0.00041 R-squared 0.29664 _____________________________________________ Regime 2: NRA > 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.27786608 0.048844640 LS 0.018122990 0.0055595105 LPRIMARY 0.089186807 0.045218543 LPOP 0.019131181 0.017608141 LRGDP0 -0.026695822 0.0050660718 NRA 0.00016556915 0.0018166993 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.0086124640 0.44350582 LS -0.0026359068 0.034757863 LPRIMARY -0.024686308 0.18163291 LPOP -0.085593732 0.060896273 LRGDP0 -0.041762350 -0.013630826 NRA -0.0040271594 0.0067633373 Observations 55.00000 Degrees of Freedom 49.00000 Sum of Squared Errors 0.00566 Residual Variance 0.00012 R-squared 0.66198 __________________________________________________ --------------------- Sensitivity specification (2a) --------------------------- -------- Baseline L_M growth regression (specification 6a in distortthr09.out) plus NRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 43.50373 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01560 LM-test for no threshold 26.08014 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.5060999 1.6607432 LM0 0.53549690 0.25373562 LS 0.22262217 0.12308138 LPRIMARY 0.41235588 0.32218998 LPOP 1.1503782 0.78296104 LRGDP0 -0.50948551 0.15593239 NRA 0.14652184 0.14046244 Observations 186.00000 Degrees of Freedom 179.00000 Sum of Squared Errors 125.34660 Residual Variance 0.70026 R-squared 0.14347 Heteroskedasticity Test (P-Value) 0.32879 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.48322 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 78.18649 Residual Variance 0.45457 Joint R-Squared 0.46573 Heteroskedasticity Test (p-value) 0.06245 _____________________________________________ Regime 1: LRGDP0 <= 8.483222 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.67407263 2.3366411 LM0 0.14204159 0.23993551 LS 0.036516678 0.095907386 LPRIMARY 0.11785290 0.29124634 LPOP 1.1787217 0.94491160 LRGDP0 0.41385909 0.18013964 NRA 1.6824154 0.33617297 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -4.7617710 5.6905425 LM0 -0.33876445 0.63090914 LS -0.15700248 0.24706952 LPRIMARY -0.48341565 0.69221243 LPOP -0.81104328 3.1933185 LRGDP0 0.060785390 0.82522844 NRA 0.98126324 2.3414002 Observations 107.00000 Degrees of Freedom 100.00000 Sum of Squared Errors 72.31652 Residual Variance 0.72317 R-squared 0.33335 _____________________________________________ Regime 2: LRGDP0 > 8.483222 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 16.174934 2.6665414 LM0 -3.9268082 0.78002037 LS 0.59957657 0.14804637 LPRIMARY -2.9859162 2.0701749 LPOP -0.074915968 0.40860991 LRGDP0 -0.029676996 0.13881413 NRA -0.081284729 0.041140083 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.6376658 21.401355 LM0 -5.4556481 0.11238112 LS 0.23972666 1.0030071 LPRIMARY -7.3620510 3.7975877 LPOP -1.0561325 1.6628556 LRGDP0 -0.66696746 0.24239869 NRA -0.17266959 0.060981574 Observations 79.00000 Degrees of Freedom 72.00000 Sum of Squared Errors 5.86997 Residual Variance 0.08153 R-squared 0.75890 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.5060999 1.6607432 LM0 0.53549690 0.25373562 LS 0.22262217 0.12308138 LPRIMARY 0.41235588 0.32218998 LPOP 1.1503782 0.78296104 LRGDP0 -0.50948551 0.15593239 NRA 0.14652184 0.14046244 Observations 186.00000 Degrees of Freedom 179.00000 Sum of Squared Errors 125.34660 Residual Variance 0.70026 R-squared 0.14347 Heteroskedasticity Test (P-Value) 0.32879 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.29590 95% Confidence Interval: [ 0.295900 , 0.360200 ] Sum of Squared Errors 95.65250 Residual Variance 0.55612 Joint R-Squared 0.34638 Heteroskedasticity Test (p-value) 0.26942 _____________________________________________ Regime 1: NRA <= 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.5711942 2.2893890 LM0 0.37501557 0.24502037 LS 0.11827975 0.10193040 LPRIMARY 0.34502485 0.30064804 LPOP 1.3690553 1.0471786 LRGDP0 -0.33687129 0.13922716 NRA 0.67141086 0.36247698 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.94912283 10.058397 LM0 -0.11887523 0.85525550 LS -0.081503836 0.32288353 LPRIMARY -0.24424531 0.94240045 LPOP -0.72862099 3.4215254 LRGDP0 -0.61548789 -0.063986066 NRA -0.039044008 1.4233941 Observations 137.00000 Degrees of Freedom 130.00000 Sum of Squared Errors 86.60879 Residual Variance 0.66622 R-squared 0.09782 _____________________________________________ Regime 2: NRA > 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 22.901368 3.7223209 LM0 -4.5613173 0.47401222 LS 0.62298842 0.25434712 LPRIMARY 0.62255948 2.6946149 LPOP -0.37087205 1.0135556 LRGDP0 -0.52838497 0.33654474 NRA 0.012695270 0.11159993 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 15.224556 30.372877 LM0 -6.1672016 -3.6322534 LS 0.11843953 1.1929324 LPRIMARY -4.8045787 7.0159978 LPOP -3.2718033 1.6156969 LRGDP0 -1.1880127 0.13124272 NRA -0.20998501 0.23665932 Observations 49.00000 Degrees of Freedom 42.00000 Sum of Squared Errors 9.04371 Residual Variance 0.21533 R-squared 0.81669 __________________________________________________ --------------------- Senstivity specification (5a) ---------------------------- ------- Baseline output growth regression (specification 5a in distortthr09.out) plus NRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.22710 LM-test for no threshold 32.86154 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.29590 LM-test for no threshold 24.67016 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.086171392 0.029332056 LS 0.013960334 0.0032710000 LPRIMARY 0.0091396332 0.0045371205 LPOP -0.064317193 0.013631550 LRGDP0 -0.012515810 0.0027645132 NRA 0.0023276259 0.0028782932 PREMIUM -3.2659300E-005 1.2348696E-005 Observations 207.00000 Degrees of Freedom 200.00000 Sum of Squared Errors 0.07196 Residual Variance 0.00036 R-squared 0.29475 Heteroskedasticity Test (P-Value) 0.00516 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.72590 95% Confidence Interval: [ 7.718818 , 7.725903 ] Sum of Squared Errors 0.05853 Residual Variance 0.00030 Joint R-Squared 0.42636 Heteroskedasticity Test (p-value) 0.02873 _____________________________________________ Regime 1: LRGDP0 <= 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.24179841 0.10316334 LS 0.012731025 0.0036753670 LPRIMARY 0.0045265593 0.0051381510 LPOP -0.11631990 0.030244933 LRGDP0 -0.0094414613 0.0075449010 NRA 0.024871165 0.012648634 PREMIUM -1.8048014E-005 8.6348279E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.44399855 -0.039598265 LS 0.0055273058 0.019934744 LPRIMARY -0.0055442167 0.014597335 LPOP -0.17559997 -0.057039834 LRGDP0 -0.024229467 0.0053465448 NRA 7.9842530E-005 0.049662487 PREMIUM -3.4972277E-005 -1.1237514E-006 Observations 88.00000 Degrees of Freedom 81.00000 Sum of Squared Errors 0.03324 Residual Variance 0.00041 R-squared 0.32541 _____________________________________________ Regime 2: LRGDP0 > 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.041137421 0.035155538 LS 0.015996642 0.0043719866 LPRIMARY -0.033337238 0.024436060 LPOP -0.043937623 0.013007490 LRGDP0 -0.020564387 0.0034814394 NRA 0.0015853738 0.0022155168 PREMIUM -0.00035102703 8.6987272E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.027767433 0.11004228 LS 0.0074275480 0.024565735 LPRIMARY -0.081231917 0.014557440 LPOP -0.069432304 -0.018442942 LRGDP0 -0.027388008 -0.013740765 NRA -0.0027570390 0.0059277867 PREMIUM -0.00052152208 -0.00018053197 Observations 119.00000 Degrees of Freedom 112.00000 Sum of Squared Errors 0.02529 Residual Variance 0.00023 R-squared 0.39145 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.086171392 0.029332056 LS 0.013960334 0.0032710000 LPRIMARY 0.0091396332 0.0045371205 LPOP -0.064317193 0.013631550 LRGDP0 -0.012515810 0.0027645132 NRA 0.0023276259 0.0028782932 PREMIUM -3.2659300E-005 1.2348696E-005 Observations 207.00000 Degrees of Freedom 200.00000 Sum of Squared Errors 0.07196 Residual Variance 0.00036 R-squared 0.29475 Heteroskedasticity Test (P-Value) 0.00516 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.10150 95% Confidence Interval: [ -0.01000 , 0.724500 ] Sum of Squared Errors 0.06222 Residual Variance 0.00032 Joint R-Squared 0.39015 Heteroskedasticity Test (p-value) 0.00376 _____________________________________________ Regime 1: NRA <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.12621456 0.055814841 LS 0.013083191 0.0032731173 LPRIMARY 0.0052314364 0.0046341872 LPOP -0.072896813 0.023643101 LRGDP0 -0.0099625415 0.0032222559 NRA 0.020751277 0.012542105 PREMIUM -2.1058884E-005 9.7587590E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.24144283 0.030782640 LS 0.0066678813 0.020821890 LPRIMARY -0.0062436616 0.016179127 LPOP -0.11923729 -0.0030900954 LRGDP0 -0.018731689 0.0034030623 NRA -0.0058352253 0.054408707 PREMIUM -5.0283472E-005 3.1451864E-006 Observations 132.00000 Degrees of Freedom 125.00000 Sum of Squared Errors 0.05131 Residual Variance 0.00041 R-squared 0.25265 _____________________________________________ Regime 2: NRA > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.12383390 0.055416804 LS 0.021223496 0.0045788046 LPRIMARY 0.030195275 0.011844861 LPOP -0.029870338 0.017268151 LRGDP0 -0.026723570 0.0038426751 NRA 0.0016410261 0.0019035737 PREMIUM -0.00037458773 6.6680067E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.057718020 0.56219805 LS -0.016051459 0.030197953 LPRIMARY -0.051878384 0.17508166 LPOP -0.080240283 0.10153679 LRGDP0 -0.039856281 -0.013216574 NRA -0.0039985671 0.0056774709 PREMIUM -0.0013118102 0.010531904 Observations 75.00000 Degrees of Freedom 68.00000 Sum of Squared Errors 0.01091 Residual Variance 0.00016 R-squared 0.54549 __________________________________________________ --------------------- Sensitivity specification (6a) --------------------------- -------- Baseline L_M growth regression (specification 6a in distortthr09.out) plus NRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 44.74404 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01560 LM-test for no threshold 26.49057 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.4523695 1.7426821 LM0 0.50255208 0.25162736 LS 0.17347574 0.12823915 LPRIMARY 0.43299587 0.31973219 LPOP 1.5194002 0.81531973 LRGDP0 -0.46956023 0.15675495 NRA 0.20345398 0.15507483 PREMIUM -0.00078312131 0.00048928616 Observations 183.00000 Degrees of Freedom 175.00000 Sum of Squared Errors 121.22280 Residual Variance 0.69270 R-squared 0.15418 Heteroskedasticity Test (P-Value) 0.44198 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 77.39118 Residual Variance 0.46342 Joint R-Squared 0.46001 Heteroskedasticity Test (p-value) 0.07516 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 1.7269332 2.7408999 LM0 0.13009876 0.24147942 LS 0.046158048 0.10217168 LPRIMARY 0.10349341 0.29239625 LPOP 1.7778914 1.1165066 LRGDP0 0.48384270 0.18026001 NRA 1.6944858 0.35424922 PREMIUM 6.1342762E-005 0.00033399205 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -3.7418029 7.4353449 LM0 -0.35749757 0.60339842 LS -0.16230621 0.26193181 LPRIMARY -0.46960323 0.69468595 LPOP -0.60295409 3.9853568 LRGDP0 0.084440096 0.83715232 NRA 0.99409817 2.4036291 PREMIUM -0.00060276358 0.00072840674 Observations 106.00000 Degrees of Freedom 98.00000 Sum of Squared Errors 72.36403 Residual Variance 0.73841 R-squared 0.34184 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 14.506933 2.4275315 LM0 -3.5688747 0.74897772 LS 0.59538569 0.15049658 LPRIMARY -0.70292775 1.4084310 LPOP -0.26050846 0.32767517 LRGDP0 -0.077055648 0.14070514 NRA -0.049339642 0.036943983 PREMIUM 0.00059456305 0.0040909207 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 9.2010883 21.652906 LM0 -5.4544677 -2.0906568 LS 0.30041239 0.98951965 LPRIMARY -6.8733819 2.0630484 LPOP -0.91428741 0.82939926 LRGDP0 -0.35283772 0.25360229 NRA -0.13364141 0.023070564 PREMIUM -0.0078356255 0.0087279891 Observations 77.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 5.02715 Residual Variance 0.07286 R-squared 0.73529 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.4523695 1.7426821 LM0 0.50255208 0.25162736 LS 0.17347574 0.12823915 LPRIMARY 0.43299587 0.31973219 LPOP 1.5194002 0.81531973 LRGDP0 -0.46956023 0.15675495 NRA 0.20345398 0.15507483 PREMIUM -0.00078312131 0.00048928616 Observations 183.00000 Degrees of Freedom 175.00000 Sum of Squared Errors 121.22280 Residual Variance 0.69270 R-squared 0.15418 Heteroskedasticity Test (P-Value) 0.44198 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.29590 95% Confidence Interval: [ 0.295900 , 0.360200 ] Sum of Squared Errors 91.11509 Residual Variance 0.54560 Joint R-Squared 0.36425 Heteroskedasticity Test (p-value) 0.40179 _____________________________________________ Regime 1: NRA <= 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 6.4686367 2.4754103 LM0 0.37420025 0.24273124 LS 0.068955018 0.10710009 LPRIMARY 0.32803835 0.29894930 LPOP 1.8242137 1.1378639 LRGDP0 -0.28243927 0.14040179 NRA 0.47541197 0.38221634 PREMIUM -0.00057443198 0.00040189198 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 1.3481681 11.320441 LM0 -0.12020625 0.84995348 LS -0.14096117 0.28758553 LPRIMARY -0.25790227 0.92572258 LPOP -0.51819916 4.0544270 LRGDP0 -0.56638059 -0.0072517689 NRA -0.27373206 1.2917477 PREMIUM -0.0013621403 0.00021327631 Observations 135.00000 Degrees of Freedom 127.00000 Sum of Squared Errors 83.90231 Residual Variance 0.66065 R-squared 0.10713 _____________________________________________ Regime 2: NRA > 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 30.740327 3.9288231 LM0 -5.5376993 0.65884910 LS 0.36264678 0.23071302 LPRIMARY 3.9870329 2.5504408 LPOP 0.23247588 0.99549038 LRGDP0 -0.61186622 0.33598463 NRA 0.055442574 0.12197022 PREMIUM -0.062350754 0.022595956 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 22.986942 38.639690 LM0 -7.5196723 -4.1785422 LS -0.094543750 0.89844665 LPRIMARY -1.0679402 9.7024101 LPOP -2.6126827 2.3403454 LRGDP0 -1.2734684 0.046663646 NRA -0.19130303 0.29465970 PREMIUM -0.10995207 -0.016216770 Observations 48.00000 Degrees of Freedom 40.00000 Sum of Squared Errors 7.21277 Residual Variance 0.18032 R-squared 0.85187 __________________________________________________ --------------------- Senstivity specification (7a) ---------------------------- ------- Baseline output growth regression (specification 5a in distortthr09.out) plus NRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 34.50660 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.29590 LM-test for no threshold 24.39617 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.068828350 0.034667443 LS 0.015646348 0.0034879411 LPRIMARY 0.0089864053 0.0046916882 LPOP -0.055216022 0.016536768 LRGDP0 -0.012129405 0.0030573787 NRA 0.0035142530 0.0028656891 PI -1.3533673E-005 2.1178845E-005 Observations 209.00000 Degrees of Freedom 202.00000 Sum of Squared Errors 0.07992 Residual Variance 0.00040 R-squared 0.26796 Heteroskedasticity Test (P-Value) 0.00101 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.24237 95% Confidence Interval: [ 6.463653 , 9.133988 ] Sum of Squared Errors 0.06620 Residual Variance 0.00034 Joint R-Squared 0.39362 Heteroskedasticity Test (p-value) 0.00903 _____________________________________________ Regime 1: LRGDP0 <= 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.085285942 0.14944414 LS 0.011945660 0.0048428875 LPRIMARY 0.0010514076 0.0070736024 LPOP -0.087911827 0.049343225 LRGDP0 -0.021968839 0.013715815 NRA 0.029852368 0.014032701 PI -9.6885531E-006 2.4353771E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.38719552 1.6303480 LS -0.013162353 0.023038443 LPRIMARY -0.052038480 0.022485926 LPOP -0.18569760 0.17423437 LRGDP0 -0.23190093 0.016388892 NRA 0.0021513846 0.17649453 PI -7.7589123E-005 6.7899025E-005 Observations 59.00000 Degrees of Freedom 52.00000 Sum of Squared Errors 0.02346 Residual Variance 0.00045 R-squared 0.24341 _____________________________________________ Regime 2: LRGDP0 > 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.0012139223 0.034171088 LS 0.022561099 0.0049180637 LPRIMARY 0.022874682 0.0079455719 LPOP -0.054177198 0.016128007 LRGDP0 -0.022228514 0.0035184502 NRA 0.0028237827 0.0025740866 PI 7.6161353E-005 8.1356156E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.13704388 0.27772830 LS -0.0095703168 0.032200504 LPRIMARY -0.053460623 0.068920856 LPOP -0.087527697 0.016039138 LRGDP0 -0.033217751 -0.00056351408 NRA -0.0022214270 0.0088351449 PI -0.00023503013 0.00032583812 Observations 150.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.04274 Residual Variance 0.00030 R-squared 0.35886 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.068828350 0.034667443 LS 0.015646348 0.0034879411 LPRIMARY 0.0089864053 0.0046916882 LPOP -0.055216022 0.016536768 LRGDP0 -0.012129405 0.0030573787 NRA 0.0035142530 0.0028656891 PI -1.3533673E-005 2.1178845E-005 Observations 209.00000 Degrees of Freedom 202.00000 Sum of Squared Errors 0.07992 Residual Variance 0.00040 R-squared 0.26796 Heteroskedasticity Test (P-Value) 0.00101 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.29590 95% Confidence Interval: [ 0.101500 , 0.482900 ] Sum of Squared Errors 0.06632 Residual Variance 0.00034 Joint R-Squared 0.39255 Heteroskedasticity Test (p-value) 0.00017 _____________________________________________ Regime 1: NRA <= 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.11367038 0.045875951 LS 0.014841803 0.0032594200 LPRIMARY 0.0066776329 0.0045233845 LPOP -0.073002151 0.018204817 LRGDP0 -0.012070209 0.0028002967 NRA 0.026316651 0.011121736 PI -6.2268593E-006 2.2915478E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.20358724 -0.010309482 LS 0.0081153895 0.021230266 LPRIMARY -0.0023071697 0.015619952 LPOP -0.10868359 -0.031185061 LRGDP0 -0.017558791 -0.0061698409 NRA 0.0026341929 0.048115253 PI -5.1790376E-005 3.8687478E-005 Observations 154.00000 Degrees of Freedom 147.00000 Sum of Squared Errors 0.06078 Residual Variance 0.00041 R-squared 0.29601 _____________________________________________ Regime 2: NRA > 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.28627730 0.042403147 LS 0.013742561 0.0056386573 LPRIMARY 0.068390645 0.043403673 LPOP 0.028601289 0.018667283 LRGDP0 -0.022208969 0.0068942108 NRA 0.00057006743 0.0017722561 PI -0.00010943543 7.7303928E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.19897185 0.36938747 LS 0.0026907932 0.025615288 LPRIMARY -0.016680555 0.15765904 LPOP -0.012769643 0.065189164 LRGDP0 -0.036491304 -0.0086963162 NRA -0.0033673318 0.0040436893 PI -0.00026450423 4.3230090E-005 Observations 55.00000 Degrees of Freedom 48.00000 Sum of Squared Errors 0.00554 Residual Variance 0.00012 R-squared 0.66903 __________________________________________________ --------------------- Senstivity specification (8a) ---------------------------- ------- Baseline L_M growth regression (specification 6a in distortthr09.out) plus NRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 40.60467 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01560 LM-test for no threshold 33.77353 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.3403169 1.6472482 LM0 0.55619625 0.26288470 LS 0.28405343 0.17565438 LPRIMARY 0.39592269 0.32552598 LPOP 1.1226575 0.77339366 LRGDP0 -0.53905251 0.16788101 NRA 0.12249565 0.14648447 PI 0.0011086835 0.0028739217 Observations 184.00000 Degrees of Freedom 176.00000 Sum of Squared Errors 124.50662 Residual Variance 0.70742 R-squared 0.14753 Heteroskedasticity Test (P-Value) 0.17242 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 72.81104 Residual Variance 0.43340 Joint R-Squared 0.50148 Heteroskedasticity Test (p-value) 0.00371 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.8073710 2.6760682 LM0 0.10924932 0.22265291 LS 0.29186629 0.15152044 LPRIMARY 0.057956306 0.28514029 LPOP 0.92531328 0.87573899 LRGDP0 0.53046106 0.17874694 NRA 1.5841300 0.31404329 PI 0.0053979700 0.0031617316 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -8.5047414 3.6703193 LM0 -0.34027911 0.54900406 LS -0.018405995 0.62289068 LPRIMARY -0.53207367 0.63528337 LPOP -1.1499280 2.6417617 LRGDP0 0.13388898 0.93786022 NRA 0.90168588 2.2120909 PI -0.00096526868 0.012384526 Observations 106.00000 Degrees of Freedom 98.00000 Sum of Squared Errors 67.69303 Residual Variance 0.69075 R-squared 0.38832 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 13.886796 2.5888523 LM0 -3.5764374 0.76207611 LS 0.56449303 0.14875234 LPRIMARY -0.79838634 1.3456991 LPOP -0.36951978 0.34871677 LRGDP0 -0.010876394 0.16350692 NRA -0.064358556 0.044980989 PI -0.0020617503 0.0019979183 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.1623431 21.371363 LM0 -5.4608412 0.12251747 LS 0.22333988 0.95156206 LPRIMARY -7.5087169 3.7664777 LPOP -1.0580454 1.6621435 LRGDP0 -0.62398099 0.32626467 NRA -0.16348057 0.076486922 PI -0.0065063911 0.0037017567 Observations 78.00000 Degrees of Freedom 70.00000 Sum of Squared Errors 5.11801 Residual Variance 0.07311 R-squared 0.73516 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.3403169 1.6472482 LM0 0.55619625 0.26288470 LS 0.28405343 0.17565438 LPRIMARY 0.39592269 0.32552598 LPOP 1.1226575 0.77339366 LRGDP0 -0.53905251 0.16788101 NRA 0.12249565 0.14648447 PI 0.0011086835 0.0028739217 Observations 184.00000 Degrees of Freedom 176.00000 Sum of Squared Errors 124.50662 Residual Variance 0.70742 R-squared 0.14753 Heteroskedasticity Test (P-Value) 0.17242 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.29590 95% Confidence Interval: [ 0.295900 , 0.360200 ] Sum of Squared Errors 95.05262 Residual Variance 0.56579 Joint R-Squared 0.34920 Heteroskedasticity Test (p-value) 0.12581 _____________________________________________ Regime 1: NRA <= 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.2795253 2.3279689 LM0 0.38253485 0.24908141 LS 0.17405445 0.16972094 LPRIMARY 0.34395950 0.30637263 LPOP 1.2937607 1.0303555 LRGDP0 -0.35783446 0.15043515 NRA 0.65233049 0.36848706 PI 0.0010940503 0.0030200208 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.58259461 9.8423443 LM0 -0.11986375 0.87073441 LS -0.15859861 0.51400426 LPRIMARY -0.25653085 0.95397577 LPOP -0.76774459 3.3132574 LRGDP0 -0.65895381 -0.062981564 NRA -0.069904145 1.4147395 PI -0.0048251905 0.0070744645 Observations 135.00000 Degrees of Freedom 127.00000 Sum of Squared Errors 86.04307 Residual Variance 0.67750 R-squared 0.10066 _____________________________________________ Regime 2: NRA > 0.295900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 22.492696 3.8207571 LM0 -4.2745365 0.64929483 LS 0.69343656 0.29214858 LPRIMARY 1.1000652 2.7633082 LPOP -0.44913409 1.0410688 LRGDP0 -0.68957163 0.45337837 NRA 0.0067623077 0.10995335 PI 0.0023779752 0.0028622709 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 14.725845 30.639218 LM0 -6.5408822 -3.0019186 LS 0.099528550 1.2773180 LPRIMARY -4.5438117 7.0322287 LPOP -3.2816344 1.5913607 LRGDP0 -1.5781932 0.22255492 NRA -0.21129306 0.23278505 PI -0.0051136641 0.0079880262 Observations 49.00000 Degrees of Freedom 41.00000 Sum of Squared Errors 9.00955 Residual Variance 0.21975 R-squared 0.81738 __________________________________________________ B) Nominal rate of assistance: 1960 -- 1990 --------------------- Senstivity specification (1b) ---------------------------- ------- Baseline output growth regression (specification 5b in distortthr09.out) plus NRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 24.21311 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.52830 LM-test for no threshold 17.85714 Bootstrap P-Value 0.02000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.038815747 0.040219460 LS 0.014430278 0.0036977412 LPRIMARY 0.016066531 0.0055178969 LPOP -0.051387009 0.018479156 LRGDP0 -0.014041911 0.0037056440 NRA 0.0065992342 0.0041072365 Observations 152.00000 Degrees of Freedom 146.00000 Sum of Squared Errors 0.05868 Residual Variance 0.00040 R-squared 0.28269 Heteroskedasticity Test (P-Value) 0.00350 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.24237 95% Confidence Interval: [ 6.505933 , 8.905376 ] Sum of Squared Errors 0.04632 Residual Variance 0.00033 Joint R-Squared 0.43381 Heteroskedasticity Test (p-value) 0.01660 _____________________________________________ Regime 1: LRGDP0 <= 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.057150343 0.15161138 LS 0.012313911 0.0044012841 LPRIMARY 0.0090384336 0.0084409711 LPOP -0.090824992 0.048285501 LRGDP0 -0.026280436 0.018236776 NRA 0.040011030 0.014229750 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.36901650 1.6540756 LS 0.0033393970 0.024441752 LPRIMARY -0.016459377 0.026559134 LPOP -0.18755342 0.078278916 LRGDP0 -0.23559547 0.012651194 NRA 0.012120721 0.16055653 Observations 45.00000 Degrees of Freedom 39.00000 Sum of Squared Errors 0.01743 Residual Variance 0.00045 R-squared 0.26725 _____________________________________________ Regime 2: LRGDP0 > 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.029838693 0.037937856 LS 0.018010437 0.0042576129 LPRIMARY 0.030166560 0.0081568866 LPOP -0.054032490 0.018143202 LRGDP0 -0.023552650 0.0037366034 NRA 0.0060005739 0.0034154866 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.13552439 0.23870756 LS -0.0023760284 0.030456811 LPRIMARY -0.052079996 0.077881328 LPOP -0.091377607 0.018625706 LRGDP0 -0.031211837 -0.0062354115 NRA -0.0026482366 0.012694928 Observations 107.00000 Degrees of Freedom 101.00000 Sum of Squared Errors 0.02889 Residual Variance 0.00029 R-squared 0.40589 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.038815747 0.040219460 LS 0.014430278 0.0036977412 LPRIMARY 0.016066531 0.0055178969 LPOP -0.051387009 0.018479156 LRGDP0 -0.014041911 0.0037056440 NRA 0.0065992342 0.0041072365 Observations 152.00000 Degrees of Freedom 146.00000 Sum of Squared Errors 0.05868 Residual Variance 0.00040 R-squared 0.28269 Heteroskedasticity Test (P-Value) 0.00350 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.13660 95% Confidence Interval: [ -0.46510 , 0.698300 ] Sum of Squared Errors 0.04676 Residual Variance 0.00033 Joint R-Squared 0.42838 Heteroskedasticity Test (p-value) 0.00070 _____________________________________________ Regime 1: NRA <= 0.136600 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.046534816 0.055370451 LS 0.012717793 0.0035411478 LPRIMARY 0.012795746 0.0051766338 LPOP -0.050003643 0.021615540 LRGDP0 -0.011831002 0.0034796177 NRA 0.033470018 0.013191702 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.0876792 0.065226488 LS 0.0057547729 0.029853315 LPRIMARY -0.029645734 0.024500502 LPOP -0.29701585 -0.0067594438 LRGDP0 -0.020701285 0.042838141 NRA -0.032144241 0.23585703 Observations 104.00000 Degrees of Freedom 98.00000 Sum of Squared Errors 0.04223 Residual Variance 0.00043 R-squared 0.24588 _____________________________________________ Regime 2: NRA > 0.136600 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.21576761 0.063985685 LS 0.027575789 0.0060118315 LPRIMARY 0.046176752 0.011853081 LPOP -0.021886552 0.023785630 LRGDP0 -0.036558084 0.0047184329 NRA 0.0027514169 0.0018416585 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.096147222 0.42009962 LS 0.0055603582 0.041546839 LPRIMARY -0.0087154292 0.15305535 LPOP -0.091819389 0.050084356 LRGDP0 -0.049474518 -0.011571321 NRA -0.0027317741 0.011950656 Observations 48.00000 Degrees of Freedom 42.00000 Sum of Squared Errors 0.00453 Residual Variance 0.00011 R-squared 0.74424 __________________________________________________ --------------------- Sensitivity specification (2b) --------------------------- -------- Baseline L_M growth regression (specification 6b in distortthr09.out) plus NRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 32.99741 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.06320 LM-test for no threshold 23.35070 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.6851335 2.1520114 LM0 0.43434974 0.35348225 LS 0.18379374 0.15689199 LPRIMARY 0.59320613 0.42612688 LPOP 0.97456029 0.98842954 LRGDP0 -0.51995343 0.19647804 NRA 0.22117490 0.22424527 Observations 128.00000 Degrees of Freedom 121.00000 Sum of Squared Errors 115.59411 Residual Variance 0.95532 R-squared 0.11322 Heteroskedasticity Test (P-Value) 0.41365 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 64.63164 Residual Variance 0.56694 Joint R-Squared 0.50418 Heteroskedasticity Test (p-value) 0.10564 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.2517183 2.6597431 LM0 0.023122962 0.31756846 LS 0.0077185343 0.12805392 LPRIMARY 0.20683525 0.38103128 LPOP 0.62615298 0.98314529 LRGDP0 0.56118545 0.21123173 NRA 1.8302139 0.32222925 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -6.4648146 5.2760709 LM0 -0.65183056 0.64555715 LS -0.24326715 0.27373398 LPRIMARY -0.53998607 1.0118981 LPOP -1.3008118 2.8992656 LRGDP0 0.076641293 0.97519964 NRA 1.1986445 2.5433468 Observations 76.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 59.78161 Residual Variance 0.86640 R-squared 0.39819 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.451156 3.3032722 LM0 -4.3135072 0.94024108 LS 0.73131139 0.21461474 LPRIMARY -2.7176081 2.0679271 LPOP -0.021463887 0.52089753 LRGDP0 -0.013431765 0.21456265 NRA -0.089813418 0.052881139 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 8.1265146 24.509326 LM0 -6.1563797 -2.0024041 LS 0.22991983 1.1519563 LPRIMARY -6.7707453 3.2716598 LPOP -1.3016062 0.99949527 LRGDP0 -0.55102186 0.40711103 NRA -0.19346045 0.024064669 Observations 52.00000 Degrees of Freedom 45.00000 Sum of Squared Errors 4.85003 Residual Variance 0.10778 R-squared 0.76610 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.6851335 2.1520114 LM0 0.43434974 0.35348225 LS 0.18379374 0.15689199 LPRIMARY 0.59320613 0.42612688 LPOP 0.97456029 0.98842954 LRGDP0 -0.51995343 0.19647804 NRA 0.22117490 0.22424527 Observations 128.00000 Degrees of Freedom 121.00000 Sum of Squared Errors 115.59411 Residual Variance 0.95532 R-squared 0.11322 Heteroskedasticity Test (P-Value) 0.41365 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.28000 95% Confidence Interval: [ 0.079200 , 0.346900 ] Sum of Squared Errors 84.91875 Residual Variance 0.74490 Joint R-Squared 0.34855 Heteroskedasticity Test (p-value) 0.33953 _____________________________________________ Regime 1: NRA <= 0.280000 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.6550724 2.8475039 LM0 0.19783732 0.33904784 LS 0.093351708 0.12806943 LPRIMARY 0.54199804 0.39496094 LPOP 1.1453504 1.2823964 LRGDP0 -0.31633013 0.16956095 NRA 1.0312944 0.46404229 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.69296847 16.412790 LM0 -0.51034465 0.97240253 LS -0.30020299 0.34775988 LPRIMARY -0.53227282 1.3242336 LPOP -1.6154670 6.5001897 LRGDP0 -0.64892615 0.37548239 NRA -0.14472983 2.1426625 Observations 95.00000 Degrees of Freedom 88.00000 Sum of Squared Errors 77.26761 Residual Variance 0.87804 R-squared 0.09424 _____________________________________________ Regime 2: NRA > 0.280000 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.759144 4.3195097 LM0 -3.6981836 0.66073620 LS 1.1953397 0.60917262 LPRIMARY -3.3093286 3.4263964 LPOP -2.1450567 1.7055615 LRGDP0 -1.2758464 0.69429870 NRA 0.18493491 0.21330665 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 6.1738778 32.874646 LM0 -7.4996060 0.11694491 LS -0.32388271 2.4015770 LPRIMARY -10.025065 7.5936299 LPOP -6.2296633 2.2198189 LRGDP0 -2.6366718 1.0056267 NRA -0.26672440 0.60301594 Observations 33.00000 Degrees of Freedom 26.00000 Sum of Squared Errors 7.65114 Residual Variance 0.29427 R-squared 0.83012 __________________________________________________ --------------------- Senstivity specification (5b) ---------------------------- ------- Baseline output growth regression (specification 5b in distortthr09.out) plus NRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.35021 LM-test for no threshold 23.07622 Bootstrap P-Value 0.00100 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.23270 LM-test for no threshold 17.73267 Bootstrap P-Value 0.03300 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.051959474 0.033078196 LS 0.011212972 0.0037071297 LPRIMARY 0.016141323 0.0052756240 LPOP -0.061617838 0.014434026 LRGDP0 -0.014586753 0.0033076607 NRA 0.0055393802 0.0041434202 PREMIUM -3.8136002E-005 1.2668078E-005 Observations 148.00000 Degrees of Freedom 141.00000 Sum of Squared Errors 0.04996 Residual Variance 0.00035 R-squared 0.32854 Heteroskedasticity Test (P-Value) 0.02349 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.66209 95% Confidence Interval: [ 7.505272 , 7.675453 ] Sum of Squared Errors 0.03949 Residual Variance 0.00029 Joint R-Squared 0.46925 Heteroskedasticity Test (p-value) 0.06912 _____________________________________________ Regime 1: LRGDP0 <= 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.14375862 0.10594539 LS 0.0092675375 0.0040430615 LPRIMARY 0.012237079 0.0062614777 LPOP -0.092356134 0.029035312 LRGDP0 -0.012510350 0.010123505 NRA 0.030355548 0.014104926 PREMIUM -2.4928342E-005 9.6300865E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.35312709 0.069148039 LS 0.0012277820 0.017473949 LPRIMARY -0.00091557686 0.024509575 LPOP -0.15517967 -0.029911199 LRGDP0 -0.036522987 0.0089501626 NRA 0.00085426169 0.058774488 PREMIUM -4.4338712E-005 -5.5329878E-006 Observations 64.00000 Degrees of Freedom 57.00000 Sum of Squared Errors 0.02293 Residual Variance 0.00040 R-squared 0.32644 _____________________________________________ Regime 2: LRGDP0 > 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.054821563 0.041999561 LS 0.016898410 0.0044161413 LPRIMARY -0.017455377 0.025905281 LPOP -0.057554445 0.013537374 LRGDP0 -0.026621618 0.0042541805 NRA 0.0026300215 0.0026768662 PREMIUM -0.00036421588 0.00012614931 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.048681086 0.13714070 LS 0.0082427725 0.028535167 LPRIMARY -0.076049566 0.033318974 LPOP -0.087977341 -0.030831964 LRGDP0 -0.034996718 -0.017298171 NRA -0.0031995689 0.0078766792 PREMIUM -0.00061146852 -9.0750182E-005 Observations 84.00000 Degrees of Freedom 77.00000 Sum of Squared Errors 0.01656 Residual Variance 0.00022 R-squared 0.48065 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.051959474 0.033078196 LS 0.011212972 0.0037071297 LPRIMARY 0.016141323 0.0052756240 LPOP -0.061617838 0.014434026 LRGDP0 -0.014586753 0.0033076607 NRA 0.0055393802 0.0041434202 PREMIUM -3.8136002E-005 1.2668078E-005 Observations 148.00000 Degrees of Freedom 141.00000 Sum of Squared Errors 0.04996 Residual Variance 0.00035 R-squared 0.32854 Heteroskedasticity Test (P-Value) 0.02349 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate -0.38110 95% Confidence Interval: [ -0.38110 , -0.38110 ] Sum of Squared Errors 0.04148 Residual Variance 0.00031 Joint R-Squared 0.44251 Heteroskedasticity Test (p-value) 0.49637 _____________________________________________ Regime 1: NRA <= -0.38110 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.91774914 0.085916288 LS 0.019241934 0.0041749441 LPRIMARY -0.012732700 0.0058462780 LPOP -0.26840930 0.024665554 LRGDP0 0.028714921 0.0040028133 NRA 0.035039563 0.071233540 PREMIUM -2.7412451E-005 1.1301220E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.0861451 -0.74935321 LS 0.011059043 0.027424824 LPRIMARY -0.024191405 -0.0012739949 LPOP -0.31675378 -0.22006481 LRGDP0 0.020869407 0.036560436 NRA -0.10457818 0.17465730 PREMIUM -4.9562841E-005 -5.2620604E-006 Observations 13.00000 Degrees of Freedom 6.00000 Sum of Squared Errors 0.00125 Residual Variance 0.00021 R-squared 0.82687 _____________________________________________ Regime 2: NRA > -0.38110 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.024229299 0.033118582 LS 0.015572210 0.0038198250 LPRIMARY 0.018975479 0.0051781096 LPOP -0.058464637 0.015292257 LRGDP0 -0.018154678 0.0030897563 NRA 0.0046793212 0.0037800307 PREMIUM -4.2605197E-005 2.0584334E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.089141720 0.040683123 LS 0.0080853530 0.023059067 LPRIMARY 0.0088263846 0.029124574 LPOP -0.088437460 -0.028491813 LRGDP0 -0.024210600 -0.012098756 NRA -0.0027295389 0.012088181 PREMIUM -8.2950491E-005 -2.2599025E-006 Observations 135.00000 Degrees of Freedom 128.00000 Sum of Squared Errors 0.04023 Residual Variance 0.00031 R-squared 0.35291 __________________________________________________ --------------------- Sensitivity specification (6b) --------------------------- -------- Baseline L_M growth regression (specification 6b in distortthr09.out) plus NRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 32.91053 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.09480 LM-test for no threshold 22.06345 Bootstrap P-Value 0.00300 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.1571472 2.4163308 LM0 0.39162697 0.34824947 LS 0.078031968 0.16911339 LPRIMARY 0.65388593 0.42279913 LPOP 1.4703861 1.0991597 LRGDP0 -0.47624330 0.20107067 NRA 0.33968058 0.27162483 PREMIUM -0.0010737116 0.00058999563 Observations 125.00000 Degrees of Freedom 117.00000 Sum of Squared Errors 110.15143 Residual Variance 0.94147 R-squared 0.13588 Heteroskedasticity Test (P-Value) 0.46950 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 64.17247 Residual Variance 0.58874 Joint R-Squared 0.49657 Heteroskedasticity Test (p-value) 0.12525 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.57501763 3.3116694 LM0 0.019374752 0.32440480 LS -0.0082470815 0.14568165 LPRIMARY 0.21702328 0.38949731 LPOP 0.86875243 1.2525112 LRGDP0 0.56266305 0.21723986 NRA 1.8265700 0.33446530 PREMIUM -0.00014418244 0.00034371932 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -7.0658898 7.4745384 LM0 -0.66673780 0.65520815 LS -0.29378311 0.28964960 LPRIMARY -0.54639144 1.0398140 LPOP -1.5861695 3.7784071 LRGDP0 0.067917456 0.98845318 NRA 1.1710180 2.5610471 PREMIUM -0.00087291080 0.00054676989 Observations 74.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 59.56372 Residual Variance 0.90248 R-squared 0.39328 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.563132 3.3112821 LM0 -4.3480979 0.97272478 LS 0.67966286 0.21674460 LPRIMARY -2.4536788 1.8566202 LPOP 0.17272039 0.52293487 LRGDP0 -0.038745932 0.23728810 NRA -0.050616099 0.049489156 PREMIUM -0.0024633737 0.0025822280 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 9.0558430 25.182975 LM0 -6.2546385 -2.0809847 LS 0.21241241 1.1044823 LPRIMARY -6.0926543 3.1306597 LPOP -1.0942343 1.1976727 LRGDP0 -0.54808374 0.42633875 NRA -0.15060028 0.056536392 PREMIUM -0.0077736581 0.0067654654 Observations 51.00000 Degrees of Freedom 43.00000 Sum of Squared Errors 4.60875 Residual Variance 0.10718 R-squared 0.77307 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.1571472 2.4163308 LM0 0.39162697 0.34824947 LS 0.078031968 0.16911339 LPRIMARY 0.65388593 0.42279913 LPOP 1.4703861 1.0991597 LRGDP0 -0.47624330 0.20107067 NRA 0.33968058 0.27162483 PREMIUM -0.0010737116 0.00058999563 Observations 125.00000 Degrees of Freedom 117.00000 Sum of Squared Errors 110.15143 Residual Variance 0.94147 R-squared 0.13588 Heteroskedasticity Test (P-Value) 0.46950 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.25550 95% Confidence Interval: [ 0.134300 , 0.346900 ] Sum of Squared Errors 79.99017 Residual Variance 0.73385 Joint R-Squared 0.37249 Heteroskedasticity Test (p-value) 0.46125 _____________________________________________ Regime 1: NRA <= 0.255500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 7.1050147 3.5369298 LM0 0.23534180 0.34194593 LS -0.0014691741 0.14472850 LPRIMARY 0.52006714 0.39889574 LPOP 1.8186419 1.6506516 LRGDP0 -0.26272083 0.18167473 NRA 0.73710212 0.48159715 PREMIUM -0.00077068160 0.00046877374 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.10964676 14.446112 LM0 -0.46079288 0.90555582 LS -0.28620579 0.29019377 LPRIMARY -0.26186204 1.3181108 LPOP -1.4166352 5.2831152 LRGDP0 -0.62204699 0.12912010 NRA -0.20682829 1.7964649 PREMIUM -0.0016894781 0.00019115064 Observations 92.00000 Degrees of Freedom 84.00000 Sum of Squared Errors 74.49024 Residual Variance 0.88679 R-squared 0.10671 _____________________________________________ Regime 2: NRA > 0.255500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 27.440781 2.6584599 LM0 -4.7851065 0.51724957 LS 0.91602184 0.50956814 LPRIMARY 0.50324010 2.0548916 LPOP -1.8074062 1.6243734 LRGDP0 -1.4706145 0.63515020 NRA 0.22830894 0.20379848 PREMIUM -0.072978346 0.022661141 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 20.948346 38.752372 LM0 -6.2447881 -3.6151122 LS -0.14360226 2.0121132 LPRIMARY -3.9861289 6.9030405 LPOP -5.7118464 2.7358652 LRGDP0 -2.7638510 -0.028291744 NRA -0.20559696 0.65400416 PREMIUM -0.11898462 -0.017906630 Observations 33.00000 Degrees of Freedom 25.00000 Sum of Squared Errors 5.49993 Residual Variance 0.22000 R-squared 0.87516 __________________________________________________ --------------------- Senstivity specification (7b) ---------------------------- ------- Baseline output growth regression (specification 5b in distortthr09.out) plus NRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 24.52664 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.10150 LM-test for no threshold 18.05515 Bootstrap P-Value 0.02400 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.033735987 0.039557412 LS 0.012883055 0.0040379314 LPRIMARY 0.015340103 0.0054510674 LPOP -0.051454102 0.018582348 LRGDP0 -0.013923513 0.0037571927 NRA 0.0069751011 0.0041163915 PI -2.2395202E-005 1.8951847E-005 Observations 150.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.05775 Residual Variance 0.00040 R-squared 0.28750 Heteroskedasticity Test (P-Value) 0.00716 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 6.67935 95% Confidence Interval: [ 6.639614 , 8.905376 ] Sum of Squared Errors 0.04559 Residual Variance 0.00034 Joint R-Squared 0.43754 Heteroskedasticity Test (p-value) 0.30052 _____________________________________________ Regime 1: LRGDP0 <= 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.73389746 0.23852604 LS 0.0064147136 0.0064454660 LPRIMARY -0.0019685446 0.0081336446 LPOP -0.077816685 0.061704264 LRGDP0 -0.14179342 0.027079953 NRA 0.10106403 0.018914191 PI -3.8840894E-005 1.8035831E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.35169393 1.4173047 LS -0.0062183997 0.021464767 LPRIMARY -0.017910488 0.028618574 LPOP -0.19875704 0.049713937 LRGDP0 -0.21999013 0.010450124 NRA 0.0087307863 0.15239151 PI -7.4191123E-005 4.1401619E-005 Observations 11.00000 Degrees of Freedom 4.00000 Sum of Squared Errors 0.00088 Residual Variance 0.00022 R-squared 0.83219 _____________________________________________ Regime 2: LRGDP0 > 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.016821637 0.039202111 LS 0.015997801 0.0035378326 LPRIMARY 0.018639366 0.0064118507 LPOP -0.050740634 0.017848361 LRGDP0 -0.015707391 0.0031674515 NRA 0.0064667060 0.0037694048 PI -0.00012532103 5.4979487E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.10041513 0.27752423 LS 0.0027559146 0.032303077 LPRIMARY -0.059302459 0.094212078 LPOP -0.091268827 0.011773467 LRGDP0 -0.042584836 -0.0085471457 NRA -0.0031226785 0.013854739 PI -0.00023749461 0.00059354714 Observations 139.00000 Degrees of Freedom 132.00000 Sum of Squared Errors 0.04470 Residual Variance 0.00034 R-squared 0.39080 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.033735987 0.039557412 LS 0.012883055 0.0040379314 LPRIMARY 0.015340103 0.0054510674 LPOP -0.051454102 0.018582348 LRGDP0 -0.013923513 0.0037571927 NRA 0.0069751011 0.0041163915 PI -2.2395202E-005 1.8951847E-005 Observations 150.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.05775 Residual Variance 0.00040 R-squared 0.28750 Heteroskedasticity Test (P-Value) 0.00716 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.13660 95% Confidence Interval: [ -0.45100 , 0.724500 ] Sum of Squared Errors 0.04626 Residual Variance 0.00034 Joint R-Squared 0.42923 Heteroskedasticity Test (p-value) 0.00134 _____________________________________________ Regime 1: NRA <= 0.136600 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.037307808 0.055641680 LS 0.011037112 0.0037625052 LPRIMARY 0.012487060 0.0052088380 LPOP -0.048187111 0.022148917 LRGDP0 -0.011612850 0.0035330779 NRA 0.033065045 0.013578248 PI -2.0692665E-005 1.8109177E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -3.2680825 0.52544908 LS -0.017209461 0.092917874 LPRIMARY -0.053907457 0.034532425 LPOP -0.71821108 0.10804295 LRGDP0 -0.030785597 0.11349529 NRA -0.47750264 0.38303779 PI -0.00086869910 0.0020824041 Observations 102.00000 Degrees of Freedom 95.00000 Sum of Squared Errors 0.04175 Residual Variance 0.00044 R-squared 0.24893 _____________________________________________ Regime 2: NRA > 0.136600 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.21674532 0.064520354 LS 0.026491822 0.0069051744 LPRIMARY 0.043780870 0.011478648 LPOP -0.021332204 0.024383824 LRGDP0 -0.035841772 0.0051290084 NRA 0.0030813663 0.0018713138 PI -3.4625796E-005 8.9308702E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.11453484 0.42578824 LS 0.0011490448 0.046186990 LPRIMARY -0.013615265 0.13907976 LPOP -0.10122553 0.055421204 LRGDP0 -0.054143297 -0.0098291443 NRA -0.0032663702 0.011991234 PI -0.00037486197 0.00031032276 Observations 48.00000 Degrees of Freedom 41.00000 Sum of Squared Errors 0.00451 Residual Variance 0.00011 R-squared 0.74499 __________________________________________________ --------------------- Senstivity specification (8b) ---------------------------- ------- Baseline L_M growth regression (specification 6b in distortthr09.out) plus NRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 30.84231 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using NRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01530 LM-test for no threshold 29.01223 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.4485000 2.1710979 LM0 0.45066483 0.35681396 LS 0.24795512 0.22083930 LPRIMARY 0.57031228 0.44595951 LPOP 0.93036951 0.97142730 LRGDP0 -0.54506251 0.20520917 NRA 0.19698609 0.22895675 PI 0.0010389717 0.0034249043 Observations 126.00000 Degrees of Freedom 118.00000 Sum of Squared Errors 114.85672 Residual Variance 0.97336 R-squared 0.11646 Heteroskedasticity Test (P-Value) 0.33855 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.474243 ] Sum of Squared Errors 57.94274 Residual Variance 0.52675 Joint R-Squared 0.55427 Heteroskedasticity Test (p-value) 0.02970 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -5.3099561 3.2634157 LM0 -0.047035338 0.28231599 LS 0.36784808 0.20672705 LPRIMARY 0.078840711 0.39607675 LPOP 0.097047686 0.88062845 LRGDP0 0.75309514 0.21430755 NRA 1.6705828 0.31037572 PI 0.0066394424 0.0036308200 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -11.706251 1.0863387 LM0 -0.60037467 0.50630400 LS -0.037336947 0.77303310 LPRIMARY -0.69746973 0.85515115 LPOP -1.6289841 1.8230794 LRGDP0 0.33305233 1.1731379 NRA 1.0622464 2.2789192 PI -0.00047696483 0.013755850 Observations 74.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 53.25411 Residual Variance 0.80688 R-squared 0.45992 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.098618 3.2706431 LM0 -4.2714460 0.92828094 LS 0.61049322 0.20335163 LPRIMARY -3.0464228 2.0059008 LPOP 0.046827345 0.52228943 LRGDP0 0.10320106 0.22393262 NRA -0.059054347 0.055900079 PI -0.0044284816 0.0025126665 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 10.688157 23.509078 LM0 -6.0908766 -2.4520153 LS 0.21192403 1.0090624 LPRIMARY -6.9779883 0.88514264 LPOP -0.97685993 1.0705146 LRGDP0 -0.33570687 0.54210899 NRA -0.16861850 0.050509808 PI -0.0093533079 0.00049634476 Observations 52.00000 Degrees of Freedom 44.00000 Sum of Squared Errors 4.68863 Residual Variance 0.10656 R-squared 0.77388 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.4485000 2.1710979 LM0 0.45066483 0.35681396 LS 0.24795512 0.22083930 LPRIMARY 0.57031228 0.44595951 LPOP 0.93036951 0.97142730 LRGDP0 -0.54506251 0.20520917 NRA 0.19698609 0.22895675 PI 0.0010389717 0.0034249043 Observations 126.00000 Degrees of Freedom 118.00000 Sum of Squared Errors 114.85672 Residual Variance 0.97336 R-squared 0.11646 Heteroskedasticity Test (P-Value) 0.33855 __________________________________________________ Threshold Estimation Threshold Variable = NRA Threshold Estimate 0.28000 95% Confidence Interval: [ -0.00660 , 0.346900 ] Sum of Squared Errors 84.37788 Residual Variance 0.76707 Joint R-Squared 0.35092 Heteroskedasticity Test (p-value) 0.26238 _____________________________________________ Regime 1: NRA <= 0.280000 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.3449559 2.9094446 LM0 0.18991364 0.33945255 LS 0.15272767 0.20793011 LPRIMARY 0.55332847 0.41955565 LPOP 1.0656764 1.2543194 LRGDP0 -0.33056313 0.17627635 NRA 1.0350925 0.47653247 PI 0.0010700957 0.0034916342 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.5583142 15.957743 LM0 -0.77968761 0.96644276 LS -0.40212698 0.56687956 LPRIMARY -0.54983666 1.3859536 LPOP -1.6508856 6.2475852 LRGDP0 -0.67639327 0.82264720 NRA -0.17727156 3.1988006 PI -0.0072096571 0.0094748447 Observations 93.00000 Degrees of Freedom 85.00000 Sum of Squared Errors 76.73398 Residual Variance 0.90275 R-squared 0.09666 _____________________________________________ Regime 2: NRA > 0.280000 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.938676 4.4922514 LM0 -3.8315662 0.74866664 LS 1.1578643 0.59358161 LPRIMARY -3.6755817 3.6090853 LPOP -2.1247173 1.6971009 LRGDP0 -1.2016121 0.65098214 NRA 0.19402852 0.22302437 PI -0.0013837032 0.0045310658 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.2019403 31.245434 LM0 -7.4305823 2.4546815 LS -0.051226356 2.3212842 LPRIMARY -10.749389 7.2235853 LPOP -6.1976517 2.2828624 LRGDP0 -2.4775371 0.92699812 NRA -0.35311963 0.63138713 PI -0.015070707 0.024374293 Observations 33.00000 Degrees of Freedom 25.00000 Sum of Squared Errors 7.64390 Residual Variance 0.30576 R-squared 0.83028 __________________________________________________ C) Relative rate of assistance: 1960 -- 2000 --------------------- Senstivity specification (1c) ---------------------------- ------- Baseline output growth regression (specification 5c in distortthr09.out) plus RRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 28.84297 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.36510 LM-test for no threshold 21.03607 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.066204983 0.037452520 LS 0.015577207 0.0034606444 LPRIMARY 0.010894990 0.0064869086 LPOP -0.054988587 0.016607576 LRGDP0 -0.012342750 0.0031155118 RRA 0.0024574667 0.0029695536 Observations 196.00000 Degrees of Freedom 190.00000 Sum of Squared Errors 0.07601 Residual Variance 0.00040 R-squared 0.24708 Heteroskedasticity Test (P-Value) 0.00033 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.19751 95% Confidence Interval: [ 6.450786 , 9.321488 ] Sum of Squared Errors 0.06439 Residual Variance 0.00035 Joint R-Squared 0.36210 Heteroskedasticity Test (p-value) 0.00427 _____________________________________________ Regime 1: LRGDP0 <= 7.197510 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.14559260 0.15483518 LS 0.010537979 0.0046983208 LPRIMARY -0.010517389 0.0095327423 LPOP -0.089371086 0.053405722 LRGDP0 -0.014272406 0.015422771 RRA 0.0083306458 0.012934463 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.68298355 2.0778479 LS -0.011148845 0.022963419 LPRIMARY -0.087435569 0.014917525 LPOP -0.25036609 0.23396605 LRGDP0 -0.26336321 0.068293339 RRA -0.079451993 0.19116326 Observations 47.00000 Degrees of Freedom 41.00000 Sum of Squared Errors 0.02141 Residual Variance 0.00052 R-squared 0.18585 _____________________________________________ Regime 2: LRGDP0 > 7.197510 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.00097526971 0.035956629 LS 0.019563834 0.0038986571 LPRIMARY 0.023221350 0.0077107057 LPOP -0.054170642 0.016497047 LRGDP0 -0.020565699 0.0030563152 RRA 0.0039044806 0.0026849861 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.13617559 0.25554026 LS -0.0083790417 0.029014682 LPRIMARY -0.051629243 0.058690582 LPOP -0.089087893 0.018588767 LRGDP0 -0.026556076 -0.0089246303 RRA -0.0044221900 0.0095585887 Observations 149.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.04298 Residual Variance 0.00030 R-squared 0.35192 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.066204983 0.037452520 LS 0.015577207 0.0034606444 LPRIMARY 0.010894990 0.0064869086 LPOP -0.054988587 0.016607576 LRGDP0 -0.012342750 0.0031155118 RRA 0.0024574667 0.0029695536 Observations 196.00000 Degrees of Freedom 190.00000 Sum of Squared Errors 0.07601 Residual Variance 0.00040 R-squared 0.24708 Heteroskedasticity Test (P-Value) 0.00033 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.11080 95% Confidence Interval: [ 0.029800 , 0.178400 ] Sum of Squared Errors 0.06528 Residual Variance 0.00035 Joint R-Squared 0.35329 Heteroskedasticity Test (p-value) 0.00043 _____________________________________________ Regime 1: RRA <= 0.110800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.10600258 0.049242425 LS 0.012547926 0.0032482875 LPRIMARY 0.0060416001 0.0067738018 LPOP -0.080759699 0.018163206 LRGDP0 -0.013973643 0.0029489425 RRA 0.033923654 0.011194438 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.20450736 -0.0032923921 LS 0.0060349347 0.018945369 LPRIMARY -0.0074309186 0.019586420 LPOP -0.11671903 -0.042264553 LRGDP0 -0.019984460 -0.0078850758 RRA 0.0094376972 0.055965289 Observations 128.00000 Degrees of Freedom 122.00000 Sum of Squared Errors 0.05146 Residual Variance 0.00042 R-squared 0.30006 _____________________________________________ Regime 2: RRA > 0.110800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.029732052 0.062035306 LS 0.042662992 0.0089483091 LPRIMARY -0.0060102619 0.018821237 LPOP -0.014372343 0.029399285 LRGDP0 -0.019399933 0.0058059795 RRA -0.0013174738 0.0025827996 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.11187971 0.21996121 LS 0.024326074 0.062395586 LPRIMARY -0.044372794 0.040711269 LPOP -0.076577552 0.061574892 LRGDP0 -0.031374219 -0.0066765465 RRA -0.0074468356 0.0037448135 Observations 68.00000 Degrees of Freedom 62.00000 Sum of Squared Errors 0.01382 Residual Variance 0.00022 R-squared 0.45457 __________________________________________________ --------------------- Sensitivity specification (2c) --------------------------- -------- Baseline L_M growth regression (specification 6c in distortthr09.out) plus RRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 42.40085 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate -0.09660 LM-test for no threshold 23.99443 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.3448182 1.7746646 LM0 0.58367712 0.26204557 LS 0.21960946 0.12720186 LPRIMARY 0.52682824 0.38301548 LPOP 1.3520897 0.80124877 LRGDP0 -0.56203341 0.16400245 RRA 0.22097531 0.15832752 Observations 174.00000 Degrees of Freedom 167.00000 Sum of Squared Errors 122.19577 Residual Variance 0.73171 R-squared 0.14827 Heteroskedasticity Test (P-Value) 0.31218 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 74.89830 Residual Variance 0.46811 Joint R-Squared 0.47794 Heteroskedasticity Test (p-value) 0.01048 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.015193840 2.3849959 LM0 0.059704628 0.25104405 LS 0.034000880 0.11025406 LPRIMARY 0.024871670 0.34818419 LPOP 0.78446923 0.89751462 LRGDP0 0.43731880 0.19489303 RRA 1.6199703 0.37557397 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -5.6914585 4.9785691 LM0 -0.44768240 0.55378854 LS -0.18859339 0.26750752 LPRIMARY -0.68042964 0.71845086 LPOP -1.2962431 2.5435979 LRGDP0 0.013934713 0.86231067 RRA 0.86085817 2.3633063 Observations 96.00000 Degrees of Freedom 89.00000 Sum of Squared Errors 69.69632 Residual Variance 0.78310 R-squared 0.34422 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 14.450512 2.6131224 LM0 -3.6093537 0.78541434 LS 0.58404140 0.15009786 LPRIMARY -0.66008046 1.4211427 LPOP -0.35808003 0.35867813 LRGDP0 -0.076883715 0.14162934 RRA -0.073138299 0.050441633 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.8646406 21.541640 LM0 -5.5172005 0.11093733 LS 0.23149978 0.97287086 LPRIMARY -7.2810359 3.8847894 LPOP -1.0679627 1.6874732 LRGDP0 -0.68156602 0.25816607 RRA -0.18380163 0.097085400 Observations 78.00000 Degrees of Freedom 71.00000 Sum of Squared Errors 5.20198 Residual Variance 0.07327 R-squared 0.73082 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.3448182 1.7746646 LM0 0.58367712 0.26204557 LS 0.21960946 0.12720186 LPRIMARY 0.52682824 0.38301548 LPOP 1.3520897 0.80124877 LRGDP0 -0.56203341 0.16400245 RRA 0.22097531 0.15832752 Observations 174.00000 Degrees of Freedom 167.00000 Sum of Squared Errors 122.19577 Residual Variance 0.73171 R-squared 0.14827 Heteroskedasticity Test (P-Value) 0.31218 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.01180 95% Confidence Interval: [ 0.011800 , 0.136900 ] Sum of Squared Errors 89.55298 Residual Variance 0.55971 Joint R-Squared 0.37579 Heteroskedasticity Test (p-value) 0.18967 _____________________________________________ Regime 1: RRA <= 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 6.0535844 3.1593997 LM0 0.25054377 0.24975014 LS 0.067044426 0.10845253 LPRIMARY 0.51467855 0.25155533 LPOP 1.7992795 1.3587396 LRGDP0 -0.15188497 0.16583720 RRA 0.78154693 0.43409715 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.13883904 12.859158 LM0 -0.25730503 0.77431855 LS -0.14552252 0.33872861 LPRIMARY 0.021630103 1.1036960 LPOP -1.0601832 4.6967546 LRGDP0 -0.62934091 0.17701018 RRA -0.20368724 1.6323774 Observations 101.00000 Degrees of Freedom 94.00000 Sum of Squared Errors 63.62871 Residual Variance 0.67690 R-squared 0.10050 _____________________________________________ Regime 2: RRA > 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 16.065541 4.2712703 LM0 -3.2363898 1.5385421 LS 1.0794757 0.42932215 LPRIMARY 1.1746278 1.6920227 LPOP 0.045111155 0.67782448 LRGDP0 -0.46318686 0.40285258 RRA -0.027516513 0.10348514 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 7.1334584 26.728260 LM0 -6.7912096 0.36049175 LS 0.23800427 2.1521617 LPRIMARY -2.3075585 4.9161455 LPOP -1.9990391 1.6935970 LRGDP0 -1.4510817 0.36274129 RRA -0.28735359 0.18862204 Observations 73.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 25.92427 Residual Variance 0.39279 R-squared 0.63790 __________________________________________________ --------------------- Senstivity specification (5c) ---------------------------- ------- Baseline output growth regression (specification 5c in distortthr09.out) plus RRP and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 8.98094 LM-test for no threshold 30.34156 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.21700 LM-test for no threshold 24.99450 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.085938273 0.031854736 LS 0.012878762 0.0033774497 LPRIMARY 0.0082184989 0.0062013155 LPOP -0.066175271 0.013733080 LRGDP0 -0.012661695 0.0028792427 RRA 0.0012281465 0.0030249992 PREMIUM -3.8787023E-005 1.4160536E-005 Observations 192.00000 Degrees of Freedom 185.00000 Sum of Squared Errors 0.06654 Residual Variance 0.00036 R-squared 0.28811 Heteroskedasticity Test (P-Value) 0.00402 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.72590 95% Confidence Interval: [ 7.197510 , 9.247973 ] Sum of Squared Errors 0.05550 Residual Variance 0.00031 Joint R-Squared 0.40623 Heteroskedasticity Test (p-value) 0.01440 _____________________________________________ Regime 1: LRGDP0 <= 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.25368776 0.11050193 LS 0.010921595 0.0037907742 LPRIMARY 0.00042819352 0.0070445356 LPOP -0.12465201 0.033422175 LRGDP0 -0.010386789 0.0080676953 RRA 0.0046384190 0.011069318 PREMIUM -3.2588842E-005 1.0766370E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.47180462 -0.037103978 LS 0.0032421966 0.019162371 LPRIMARY -0.013379096 0.015606228 LPOP -0.19015948 -0.057821163 LRGDP0 -0.026199472 0.0065549504 RRA -0.017104838 0.026505103 PREMIUM -5.6991283E-005 -9.4367213E-006 Observations 74.00000 Degrees of Freedom 67.00000 Sum of Squared Errors 0.03026 Residual Variance 0.00045 R-squared 0.31262 _____________________________________________ Regime 2: LRGDP0 > 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.043257758 0.037256557 LS 0.015629363 0.0045413633 LPRIMARY -0.033678026 0.024463225 LPOP -0.043901749 0.013261275 LRGDP0 -0.020667536 0.0036174386 RRA 0.0021350543 0.0024274333 PREMIUM -0.00034842490 8.6767384E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.048832675 0.42005507 LS -0.015153371 0.028991085 LPRIMARY -0.087290657 0.057289785 LPOP -0.070882933 0.033224426 LRGDP0 -0.033143278 -0.013071565 RRA -0.0032238831 0.0069742264 PREMIUM -0.0019076808 -0.00016225468 Observations 118.00000 Degrees of Freedom 111.00000 Sum of Squared Errors 0.02524 Residual Variance 0.00023 R-squared 0.38831 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.085938273 0.031854736 LS 0.012878762 0.0033774497 LPRIMARY 0.0082184989 0.0062013155 LPOP -0.066175271 0.013733080 LRGDP0 -0.012661695 0.0028792427 RRA 0.0012281465 0.0030249992 PREMIUM -3.8787023E-005 1.4160536E-005 Observations 192.00000 Degrees of Freedom 185.00000 Sum of Squared Errors 0.06654 Residual Variance 0.00036 R-squared 0.28811 Heteroskedasticity Test (P-Value) 0.00402 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.17840 95% Confidence Interval: [ 0.155200 , 0.262900 ] Sum of Squared Errors 0.05438 Residual Variance 0.00031 Joint R-Squared 0.41814 Heteroskedasticity Test (p-value) 0.00087 _____________________________________________ Regime 1: RRA <= 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.12747461 0.045198188 LS 0.010670745 0.0032868385 LPRIMARY 0.0042561021 0.0065838936 LPOP -0.085024629 0.018259446 LRGDP0 -0.012519723 0.0030016202 RRA 0.018896038 0.010012425 PREMIUM -2.8802163E-005 9.9967381E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.21798866 -0.026511274 LS 0.0040371492 0.017122336 LPRIMARY -0.0086483293 0.017978394 LPOP -0.12128384 -0.042824353 LRGDP0 -0.018402898 -0.0061943927 RRA -0.0025482671 0.038520390 PREMIUM -5.0390358E-005 -9.2085559E-006 Observations 132.00000 Degrees of Freedom 125.00000 Sum of Squared Errors 0.04873 Residual Variance 0.00039 R-squared 0.28682 _____________________________________________ Regime 2: RRA > 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.35175836 0.070904688 LS 0.018746326 0.0047580619 LPRIMARY 0.049914131 0.012463733 LPOP 0.035696431 0.021462671 LRGDP0 -0.029722541 0.0042823434 RRA 0.0012679894 0.0017682907 PREMIUM -0.00067433535 9.9243712E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.20235235 0.51226642 LS 0.0052112401 0.028072127 LPRIMARY 0.025485214 0.093032123 LPOP -0.019913529 0.081389267 LRGDP0 -0.040312444 -0.021315977 RRA -0.0021978605 0.0056582437 PREMIUM -0.00093571718 -0.00047756309 Observations 60.00000 Degrees of Freedom 53.00000 Sum of Squared Errors 0.00565 Residual Variance 0.00011 R-squared 0.74935 __________________________________________________ --------------------- Sensitivity specification (6c) --------------------------- -------- Baseline L_M growth regression (specification 6c in distortthr09.out) plus RRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 43.12342 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01180 LM-test for no threshold 29.16697 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.3973557 1.8756210 LM0 0.54365199 0.26212077 LS 0.16792737 0.13494573 LPRIMARY 0.53543072 0.37669739 LPOP 1.7389394 0.83971431 LRGDP0 -0.52370287 0.16574504 RRA 0.29012577 0.17461457 PREMIUM -0.00079910300 0.00055379204 Observations 171.00000 Degrees of Freedom 163.00000 Sum of Squared Errors 117.62625 Residual Variance 0.72163 R-squared 0.16264 Heteroskedasticity Test (P-Value) 0.43253 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 74.30688 Residual Variance 0.47940 Joint R-Squared 0.47102 Heteroskedasticity Test (p-value) 0.01499 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.78047203 2.7967836 LM0 0.055098417 0.25744319 LS 0.031867562 0.11854479 LPRIMARY 0.030396091 0.35118432 LPOP 1.1218699 1.0760283 LRGDP0 0.45457162 0.19636741 RRA 1.6110173 0.39061114 PREMIUM -2.6100232E-005 0.00037672709 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -6.0143574 6.5756166 LM0 -0.46517877 0.56417466 LS -0.20920589 0.28172621 LPRIMARY -0.68330972 0.72946964 LPOP -1.4245350 3.2347982 LRGDP0 0.028397653 0.87988301 RRA 0.82492359 2.3819533 PREMIUM -0.00077226721 0.00071839373 Observations 94.00000 Degrees of Freedom 86.00000 Sum of Squared Errors 69.25143 Residual Variance 0.80525 R-squared 0.34115 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 14.622468 2.4137724 LM0 -3.5812903 0.76146555 LS 0.57921241 0.14652143 LPRIMARY -0.62363704 1.4132273 LPOP -0.25137524 0.33184544 LRGDP0 -0.076231279 0.14501660 RRA -0.038573826 0.044693736 PREMIUM 0.00066136805 0.0040845812 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 6.5469287 21.792220 LM0 -5.4934943 0.10436513 LS 0.22195248 0.96128251 LPRIMARY -6.9294090 3.8552299 LPOP -0.91573020 1.8406318 LRGDP0 -0.70274218 0.26175479 RRA -0.13852263 0.12967432 PREMIUM -0.0082057123 0.0087650193 Observations 77.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 5.05545 Residual Variance 0.07327 R-squared 0.73380 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.3973557 1.8756210 LM0 0.54365199 0.26212077 LS 0.16792737 0.13494573 LPRIMARY 0.53543072 0.37669739 LPOP 1.7389394 0.83971431 LRGDP0 -0.52370287 0.16574504 RRA 0.29012577 0.17461457 PREMIUM -0.00079910300 0.00055379204 Observations 171.00000 Degrees of Freedom 163.00000 Sum of Squared Errors 117.62625 Residual Variance 0.72163 R-squared 0.16264 Heteroskedasticity Test (P-Value) 0.43253 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.13690 95% Confidence Interval: [ -0.09660 , 0.178400 ] Sum of Squared Errors 88.33785 Residual Variance 0.56992 Joint R-Squared 0.37114 Heteroskedasticity Test (p-value) 0.15647 _____________________________________________ Regime 1: RRA <= 0.136900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 6.9652973 3.1645038 LM0 0.27947340 0.25017066 LS 0.062992801 0.11963488 LPRIMARY 0.59454834 0.25946980 LPOP 1.9153360 1.3907546 LRGDP0 -0.25721989 0.15628270 RRA 0.34836387 0.43196456 PREMIUM -0.00069167582 0.00046127209 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.60117084 13.185778 LM0 -0.26850355 0.76996924 LS -0.19607115 0.29830713 LPRIMARY 0.016919270 1.1048084 LPOP -1.0552322 4.7523138 LRGDP0 -0.56409680 0.18285057 RRA -0.53387634 1.5438084 PREMIUM -0.0015983327 0.00026312001 Observations 113.00000 Degrees of Freedom 105.00000 Sum of Squared Errors 66.43537 Residual Variance 0.63272 R-squared 0.10817 _____________________________________________ Regime 2: RRA > 0.136900 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.166613 5.9494205 LM0 -3.7069793 1.4589351 LS 1.3273910 0.58223979 LPRIMARY 1.3854022 1.8906412 LPOP -0.026436619 0.85859419 LRGDP0 -0.46315271 0.45985962 RRA -0.023807507 0.14917441 PREMIUM 0.0021438679 0.0085027395 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 4.6292197 28.827477 LM0 -6.5664920 0.072670353 LS 0.15832312 2.4957594 LPRIMARY -2.6552961 5.0910590 LPOP -2.0650871 2.1013250 LRGDP0 -1.4382572 0.44213867 RRA -0.31618934 0.30803598 PREMIUM -0.014521501 0.019187943 Observations 58.00000 Degrees of Freedom 50.00000 Sum of Squared Errors 21.90248 Residual Variance 0.43805 R-squared 0.66767 __________________________________________________ --------------------- Senstivity specification (7c) ---------------------------- ------- Baseline output growth regression (specification 5c in distortthr09.out) plus RRP and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 29.81230 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.21700 LM-test for no threshold 26.07523 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.064651023 0.037024719 LS 0.014832544 0.0036994916 LPRIMARY 0.0085717298 0.0066103060 LPOP -0.055139013 0.016650175 LRGDP0 -0.012187620 0.0031882095 RRA 0.0028322704 0.0029585522 PI -1.6051264E-005 2.0977943E-005 Observations 194.00000 Degrees of Freedom 187.00000 Sum of Squared Errors 0.07509 Residual Variance 0.00040 R-squared 0.25081 Heteroskedasticity Test (P-Value) 0.00068 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 6.67935 95% Confidence Interval: [ 6.463653 , 8.475141 ] Sum of Squared Errors 0.06264 Residual Variance 0.00035 Joint R-Squared 0.37500 Heteroskedasticity Test (p-value) 0.06033 _____________________________________________ Regime 1: LRGDP0 <= 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.42888044 0.30468172 LS -0.0080947126 0.0067407503 LPRIMARY -0.022973497 0.014125954 LPOP 0.0083320992 0.092995784 LRGDP0 -0.058642874 0.048885409 RRA 0.032457333 0.027512504 PI -5.8030321E-005 2.1131134E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.42909720 1.7809560 LS -0.023478523 0.018775487 LPRIMARY -0.073546713 0.0047976557 LPOP -0.20872249 0.25609097 LRGDP0 -0.21304041 0.054658854 RRA -0.090856422 0.15079233 PI -9.9447343E-005 1.8711597E-005 Observations 11.00000 Degrees of Freedom 4.00000 Sum of Squared Errors 0.00288 Residual Variance 0.00072 R-squared 0.50413 _____________________________________________ Regime 2: LRGDP0 > 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.050392221 0.035622075 LS 0.019105913 0.0032273151 LPRIMARY 0.0082160249 0.0061299479 LPOP -0.053811172 0.016063514 LRGDP0 -0.014052187 0.0027379174 RRA 0.0034410950 0.0028886492 PI -0.00010121131 4.7224222E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.12334245 0.069012892 LS 0.011004689 0.032332704 LPRIMARY -0.0037986731 0.038690158 LPOP -0.087062631 -0.022326685 LRGDP0 -0.029391785 -0.0083296865 RRA -0.0024850305 0.0092979082 PI -0.00020132106 0.00023867147 Observations 183.00000 Degrees of Freedom 176.00000 Sum of Squared Errors 0.05976 Residual Variance 0.00034 R-squared 0.35141 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.064651023 0.037024719 LS 0.014832544 0.0036994916 LPRIMARY 0.0085717298 0.0066103060 LPOP -0.055139013 0.016650175 LRGDP0 -0.012187620 0.0031882095 RRA 0.0028322704 0.0029585522 PI -1.6051264E-005 2.0977943E-005 Observations 194.00000 Degrees of Freedom 187.00000 Sum of Squared Errors 0.07509 Residual Variance 0.00040 R-squared 0.25081 Heteroskedasticity Test (P-Value) 0.00068 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.17840 95% Confidence Interval: [ 0.011800 , 0.262900 ] Sum of Squared Errors 0.06162 Residual Variance 0.00034 Joint R-Squared 0.38525 Heteroskedasticity Test (p-value) 0.00044 _____________________________________________ Regime 1: RRA <= 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.10878513 0.046643330 LS 0.012722757 0.0034545934 LPRIMARY 0.0039738057 0.0069146334 LPOP -0.081441339 0.017598220 LRGDP0 -0.014087296 0.0030403179 RRA 0.029966338 0.010285345 PI -3.2005242E-006 2.4649252E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.20857132 0.011771243 LS 0.0057480825 0.019797629 LPRIMARY -0.0099363591 0.018888254 LPOP -0.11902956 -0.033245704 LRGDP0 -0.020880398 -0.0074130640 RRA 0.0098070616 0.056866549 PI -5.1513058E-005 4.6897709E-005 Observations 132.00000 Degrees of Freedom 125.00000 Sum of Squared Errors 0.05206 Residual Variance 0.00042 R-squared 0.29484 _____________________________________________ Regime 2: RRA > 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.14959298 0.065078408 LS 0.031652228 0.0099332049 LPRIMARY -0.016356319 0.015101621 LPOP 0.026613996 0.023631637 LRGDP0 -0.013396064 0.0047417954 RRA -4.8181714E-005 0.0023820592 PI -0.00025556190 3.4016166E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.069249786 0.27714666 LS 0.0081230073 0.052129432 LPRIMARY -0.050981960 0.025245043 LPOP -0.066445732 0.072932005 LRGDP0 -0.027134322 -0.0041021445 RRA -0.0055930660 0.0060701482 PI -0.00034686833 -8.2607373E-005 Observations 62.00000 Degrees of Freedom 55.00000 Sum of Squared Errors 0.00956 Residual Variance 0.00017 R-squared 0.60440 __________________________________________________ --------------------- Senstivity specification (8c) ---------------------------- ------- Baseline L_M growth regression (specification 6c in distortthr09.out) plus RRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 39.42680 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate -0.12530 LM-test for no threshold 29.81740 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.1832618 1.7556851 LM0 0.60623937 0.27314630 LS 0.27678472 0.17395400 LPRIMARY 0.51374379 0.38326899 LPOP 1.3280577 0.79405135 LRGDP0 -0.59001234 0.17354557 NRP 0.19710386 0.15810812 PI 0.0010377543 0.0027859876 Observations 172.00000 Degrees of Freedom 164.00000 Sum of Squared Errors 121.38384 Residual Variance 0.74015 R-squared 0.15216 Heteroskedasticity Test (P-Value) 0.18646 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 71.30551 Residual Variance 0.45709 Joint R-Squared 0.50195 Heteroskedasticity Test (p-value) 0.00849 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -2.3915697 2.8063947 LM0 0.065827693 0.24496457 LS 0.25339153 0.15315095 LPRIMARY -0.032827798 0.34042463 LPOP 0.42821804 0.88079597 LRGDP0 0.50387646 0.19920794 NRP 1.4324408 0.34666227 PI 0.0045933026 0.0027603359 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -9.6008292 3.3135917 LM0 -0.42903469 0.54595825 LS -0.059078849 0.59857088 LPRIMARY -0.73293853 0.64426726 LPOP -1.7120532 2.1545781 LRGDP0 0.071369394 0.97735304 NRP 0.70319236 2.1218657 PI -0.00096324291 0.010986691 Observations 94.00000 Degrees of Freedom 86.00000 Sum of Squared Errors 66.15800 Residual Variance 0.76928 R-squared 0.37328 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 13.937818 2.5808988 LM0 -3.6015451 0.77503901 LS 0.54444519 0.14442658 LPRIMARY -0.73410205 1.3495459 LPOP -0.37386214 0.35085469 LRGDP0 0.0019923164 0.16504832 NRP -0.058231126 0.053097447 PI -0.0022703639 0.0020047668 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.3129227 21.479896 LM0 -5.5185105 0.12851201 LS 0.21374262 0.91961157 LPRIMARY -7.4582575 3.8419234 LPOP -1.0687466 1.6862614 LRGDP0 -0.62697271 0.34271740 NRP -0.17332123 0.11521409 PI -0.0067102742 0.0034394602 Observations 78.00000 Degrees of Freedom 70.00000 Sum of Squared Errors 5.14751 Residual Variance 0.07354 R-squared 0.73364 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.1832618 1.7556851 LM0 0.60623937 0.27314630 LS 0.27678472 0.17395400 LPRIMARY 0.51374379 0.38326899 LPOP 1.3280577 0.79405135 LRGDP0 -0.59001234 0.17354557 NRP 0.19710386 0.15810812 PI 0.0010377543 0.0027859876 Observations 172.00000 Degrees of Freedom 164.00000 Sum of Squared Errors 121.38384 Residual Variance 0.74015 R-squared 0.15216 Heteroskedasticity Test (P-Value) 0.18646 __________________________________________________ Threshold Estimation Threshold Variable = NRP Threshold Estimate 0.01180 95% Confidence Interval: [ 0.011800 , 0.136900 ] Sum of Squared Errors 86.65279 Residual Variance 0.55547 Joint R-Squared 0.39475 Heteroskedasticity Test (p-value) 0.23089 _____________________________________________ Regime 1: NRP <= 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 6.8717525 3.2749661 LM0 0.21851469 0.24722356 LS -0.079388199 0.13367357 LPRIMARY 0.54601100 0.26215375 LPOP 1.9808820 1.4019401 LRGDP0 -0.10132437 0.16491087 NRP 0.71568713 0.43091993 PI -0.0033755759 0.0018877525 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 0.45281892 13.876136 LM0 -0.28609136 0.70307286 LS -0.34322767 0.18934049 LPRIMARY 0.032189646 1.0869844 LPOP -0.76692057 4.9619288 LRGDP0 -0.43576654 0.22741674 NRP -0.25328642 1.5602902 PI -0.0073854768 0.00032441898 Observations 99.00000 Degrees of Freedom 91.00000 Sum of Squared Errors 61.94176 Residual Variance 0.68068 R-squared 0.11977 _____________________________________________ Regime 2: NRP > 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 14.906383 4.2352997 LM0 -2.9647249 1.4808050 LS 1.3021033 0.45799451 LPRIMARY 1.3683312 1.5499467 LPOP 0.053724993 0.61363002 LRGDP0 -0.57988772 0.40956456 NRP -0.096949672 0.12932581 PI 0.0046050143 0.0047242049 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 6.2364849 23.338121 LM0 -5.8671028 0.44694010 LS 0.40269261 2.2465450 LPRIMARY -1.8657884 4.4062267 LPOP -1.2253683 1.5454351 LRGDP0 -1.5285869 0.22285883 NRP -0.35120474 0.16821111 PI -0.0051213037 0.013864456 Observations 73.00000 Degrees of Freedom 65.00000 Sum of Squared Errors 24.71103 Residual Variance 0.38017 R-squared 0.65485 __________________________________________________ D) Relative rate of assistance: 1960 -- 1990 --------------------- Senstivity specification (1d) ---------------------------- ------- Baseline output growth regression (specification 5d in distortthr09.out) plus RRP as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 21.39835 Bootstrap P-Value 0.00300 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.33590 LM-test for no threshold 18.82769 Bootstrap P-Value 0.00400 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.028380848 0.043560295 LS 0.013327818 0.0038553249 LPRIMARY 0.018964135 0.0075738259 LPOP -0.050283114 0.018639689 LRGDP0 -0.014354722 0.0038703759 RRA 0.0063249630 0.0040801396 Observations 142.00000 Degrees of Freedom 136.00000 Sum of Squared Errors 0.05592 Residual Variance 0.00041 R-squared 0.26164 Heteroskedasticity Test (P-Value) 0.00302 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.19676 95% Confidence Interval: [ 6.679347 , 8.980939 ] Sum of Squared Errors 0.04586 Residual Variance 0.00035 Joint R-Squared 0.39447 Heteroskedasticity Test (p-value) 0.01009 _____________________________________________ Regime 1: LRGDP0 <= 7.196761 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.084689221 0.18110754 LS 0.010455000 0.0047975268 LPRIMARY -0.00031932529 0.013250003 LPOP -0.096034219 0.056723732 LRGDP0 -0.024099671 0.021894606 RRA 0.019133349 0.013720141 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.43965999 1.6621418 LS -0.0020153842 0.023086582 LPRIMARY -0.075058308 0.027523932 LPOP -0.31301887 0.085839332 LRGDP0 -0.24998078 0.020337207 RRA -0.0093574302 0.11185719 Observations 36.00000 Degrees of Freedom 30.00000 Sum of Squared Errors 0.01698 Residual Variance 0.00057 R-squared 0.18886 _____________________________________________ Regime 2: LRGDP0 > 7.196761 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.033440276 0.039484530 LS 0.017790171 0.0043848890 LPRIMARY 0.030642814 0.0082478385 LPOP -0.054168746 0.018250772 LRGDP0 -0.023847148 0.0038253986 RRA 0.0064694644 0.0035221310 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.13596341 0.24806538 LS -0.0021827275 0.031069463 LPRIMARY -0.050576538 0.076069875 LPOP -0.091274272 0.022782915 LRGDP0 -0.031832452 -0.0062569042 RRA -0.0026359957 0.014199992 Observations 106.00000 Degrees of Freedom 100.00000 Sum of Squared Errors 0.02888 Residual Variance 0.00029 R-squared 0.40260 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.028380848 0.043560295 LS 0.013327818 0.0038553249 LPRIMARY 0.018964135 0.0075738259 LPOP -0.050283114 0.018639689 LRGDP0 -0.014354722 0.0038703759 RRA 0.0063249630 0.0040801396 Observations 142.00000 Degrees of Freedom 136.00000 Sum of Squared Errors 0.05592 Residual Variance 0.00041 R-squared 0.26164 Heteroskedasticity Test (P-Value) 0.00302 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate -0.54490 95% Confidence Interval: [ -0.58450 , -0.44770 ] Sum of Squared Errors 0.04531 Residual Variance 0.00035 Joint R-Squared 0.40174 Heteroskedasticity Test (p-value) 0.06868 _____________________________________________ Regime 1: RRA <= -0.54490 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.78084620 0.20705035 LS 0.024118931 0.0049398729 LPRIMARY -0.028657876 0.011536573 LPOP -0.28428811 0.080344421 LRGDP0 0.017304392 0.0069073155 RRA 0.19597062 0.054216301 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.4592779 -0.25305542 LS 0.011278449 0.033801082 LPRIMARY -0.061537414 -0.0043787417 LPOP -0.53915298 -0.086271443 LRGDP0 0.0037660536 0.041141365 RRA 0.043585261 0.37817825 Observations 16.00000 Degrees of Freedom 10.00000 Sum of Squared Errors 0.00403 Residual Variance 0.00040 R-squared 0.63606 _____________________________________________ Regime 2: RRA > -0.54490 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.0083060728 0.043737296 LS 0.012796384 0.0041795628 LPRIMARY 0.026668710 0.0073542082 LPOP -0.056242849 0.019071470 LRGDP0 -0.018143901 0.0033030637 RRA 0.0047184380 0.0036812573 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.095974389 0.077419027 LS 0.0041935634 0.023184116 LPRIMARY 0.0092967179 0.042943151 LPOP -0.097462189 -0.016629222 LRGDP0 -0.025669207 -0.011295465 RRA -0.0027062117 0.011976778 Observations 126.00000 Degrees of Freedom 120.00000 Sum of Squared Errors 0.04128 Residual Variance 0.00034 R-squared 0.28693 __________________________________________________ --------------------- Sensitivity specification (2d) --------------------------- -------- Baseline L_M growth regression (specification 6d in distortthr09.out) plus RRA as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 8.19420 LM-test for no threshold 32.44929 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate -0.27870 LM-test for no threshold 22.83444 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.0064645 2.3686066 LM0 0.53770905 0.35785688 LS 0.15545294 0.16416438 LPRIMARY 0.72045011 0.46434946 LPOP 1.2832886 1.0217753 LRGDP0 -0.61099503 0.20629284 RRA 0.34834859 0.25673529 Observations 121.00000 Degrees of Freedom 114.00000 Sum of Squared Errors 112.01887 Residual Variance 0.98262 R-squared 0.12715 Heteroskedasticity Test (P-Value) 0.41146 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 61.26748 Residual Variance 0.57259 Joint R-Squared 0.52261 Heteroskedasticity Test (p-value) 0.01736 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.4298612 2.7947926 LM0 0.050520590 0.31905667 LS -0.041306661 0.15791727 LPRIMARY 0.18433411 0.40630606 LPOP 0.068124853 0.91269594 LRGDP0 0.43793105 0.23832045 RRA 1.8041608 0.41221150 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -6.9076547 4.0479323 LM0 -0.57483047 0.67587165 LS -0.35082451 0.26821118 LPRIMARY -0.61202577 0.98069399 LPOP -1.7207592 1.8570089 LRGDP0 -0.029177029 0.90503913 RRA 0.99622626 2.6120953 Observations 69.00000 Degrees of Freedom 62.00000 Sum of Squared Errors 56.38089 Residual Variance 0.90937 R-squared 0.40933 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.627744 3.2949486 LM0 -4.3662199 0.95312652 LS 0.70720768 0.21143721 LPRIMARY -2.6615497 2.0719038 LPOP -0.023335344 0.53342151 LRGDP0 -0.00088060209 0.22110738 RRA -0.083938934 0.064448620 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 11.169645 24.085843 LM0 -6.2343479 -2.4980920 LS 0.29279075 1.1216246 LPRIMARY -6.7224812 1.3993818 LPOP -1.0688415 1.0221708 LRGDP0 -0.43425106 0.43248986 RRA -0.21025823 0.042380360 Observations 52.00000 Degrees of Freedom 45.00000 Sum of Squared Errors 4.88658 Residual Variance 0.10859 R-squared 0.76434 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 7.0064645 2.3686066 LM0 0.53770905 0.35785688 LS 0.15545294 0.16416438 LPRIMARY 0.72045011 0.46434946 LPOP 1.2832886 1.0217753 LRGDP0 -0.61099503 0.20629284 RRA 0.34834859 0.25673529 Observations 121.00000 Degrees of Freedom 114.00000 Sum of Squared Errors 112.01887 Residual Variance 0.98262 R-squared 0.12715 Heteroskedasticity Test (P-Value) 0.41146 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.01180 95% Confidence Interval: [ 0.011800 , 0.110800 ] Sum of Squared Errors 74.62609 Residual Variance 0.69744 Joint R-Squared 0.41852 Heteroskedasticity Test (p-value) 0.36217 _____________________________________________ Regime 1: RRA <= 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.3839381 3.9839043 LM0 0.12629864 0.31957046 LS -0.023991382 0.14096592 LPRIMARY 0.68321042 0.31173398 LPOP 1.1539665 1.6929779 LRGDP0 -0.16119704 0.19715120 RRA 1.3015317 0.56857071 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.4245143 14.400701 LM0 -0.53054723 0.79364058 LS -0.30028458 0.33326221 LPRIMARY 0.072211819 1.4372441 LPOP -2.1642701 5.0006707 LRGDP0 -0.72215389 0.23206477 RRA -0.011811541 2.4159303 Observations 73.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 54.64604 Residual Variance 0.82797 R-squared 0.13605 _____________________________________________ Regime 2: RRA > 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 20.149770 4.6682119 LM0 -4.1853540 1.8115899 LS 1.6218346 0.59356453 LPRIMARY 1.4878883 1.7418874 LPOP 0.86598667 0.82423172 LRGDP0 -0.38051310 0.54302871 RRA -0.11554350 0.15459886 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 10.396303 29.999387 LM0 -8.2665098 -0.040396763 LS 0.44233782 3.2054756 LPRIMARY -2.1562920 5.0246653 LPOP -1.8290313 3.1205244 LRGDP0 -1.8391893 0.80399773 RRA -0.44536676 0.26248150 Observations 48.00000 Degrees of Freedom 41.00000 Sum of Squared Errors 19.98005 Residual Variance 0.48732 R-squared 0.69260 __________________________________________________ --------------------- Senstivity specification (5d) ---------------------------- ------- Baseline output growth regression (specification 5d in distortthr09.out) plus RRP and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.35021 LM-test for no threshold 21.41543 Bootstrap P-Value 0.00300 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.35950 LM-test for no threshold 17.16713 Bootstrap P-Value 0.02800 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.046717485 0.036353113 LS 0.0096744885 0.0038679236 LPRIMARY 0.017123310 0.0071734463 LPOP -0.063270409 0.014638195 LRGDP0 -0.015037430 0.0034494648 RRA 0.0050986279 0.0041346193 PREMIUM -4.3182871E-005 1.4035691E-005 Observations 138.00000 Degrees of Freedom 131.00000 Sum of Squared Errors 0.04660 Residual Variance 0.00036 R-squared 0.32088 Heteroskedasticity Test (P-Value) 0.02344 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.66209 95% Confidence Interval: [ 7.196761 , 9.247973 ] Sum of Squared Errors 0.03804 Residual Variance 0.00031 Joint R-Squared 0.44564 Heteroskedasticity Test (p-value) 0.04702 _____________________________________________ Regime 1: LRGDP0 <= 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.13920863 0.11393504 LS 0.0071739819 0.0042414070 LPRIMARY 0.011196851 0.0087428629 LPOP -0.10052462 0.033405203 LRGDP0 -0.015181806 0.010398171 RRA 0.014096800 0.012290855 PREMIUM -3.6534780E-005 1.1547219E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.36312129 0.11673415 LS -0.0022283123 0.016999667 LPRIMARY -0.0071570650 0.028651030 LPOP -0.18297578 -0.030085477 LRGDP0 -0.046010528 0.0065777287 RRA -0.010529860 0.038340237 PREMIUM -6.1045964E-005 -1.2403961E-005 Observations 55.00000 Degrees of Freedom 48.00000 Sum of Squared Errors 0.02152 Residual Variance 0.00045 R-squared 0.30184 _____________________________________________ Regime 2: LRGDP0 > 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.058855037 0.045052701 LS 0.016670953 0.0046301222 LPRIMARY -0.018128264 0.026158072 LPOP -0.056879808 0.013959006 LRGDP0 -0.026773013 0.0044618084 RRA 0.0033486413 0.0029126547 PREMIUM -0.00036047059 0.00012585281 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.061849997 0.36741816 LS -0.0074689696 0.028624737 LPRIMARY -0.076868402 0.057906930 LPOP -0.091642508 0.020491131 LRGDP0 -0.035590153 -0.014355073 RRA -0.0033087535 0.0090574445 PREMIUM -0.0018171385 -8.7471240E-005 Observations 83.00000 Degrees of Freedom 76.00000 Sum of Squared Errors 0.01652 Residual Variance 0.00022 R-squared 0.47775 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.046717485 0.036353113 LS 0.0096744885 0.0038679236 LPRIMARY 0.017123310 0.0071734463 LPOP -0.063270409 0.014638195 LRGDP0 -0.015037430 0.0034494648 RRA 0.0050986279 0.0041346193 PREMIUM -4.3182871E-005 1.4035691E-005 Observations 138.00000 Degrees of Freedom 131.00000 Sum of Squared Errors 0.04660 Residual Variance 0.00036 R-squared 0.32088 Heteroskedasticity Test (P-Value) 0.02344 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.17840 95% Confidence Interval: [ -0.62320 , 0.442800 ] Sum of Squared Errors 0.03887 Residual Variance 0.00031 Joint R-Squared 0.43347 Heteroskedasticity Test (p-value) 0.01270 _____________________________________________ Regime 1: RRA <= 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.070745048 0.046411821 LS 0.0080277942 0.0038919789 LPRIMARY 0.013448252 0.0076711840 LPOP -0.074708642 0.017855482 LRGDP0 -0.014831280 0.0034768465 RRA 0.021323991 0.011086301 PREMIUM -3.5892721E-005 1.1670179E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.8909996 0.27984033 LS -0.0016346245 0.033697489 LPRIMARY -0.051022942 0.029795061 LPOP -0.74710016 0.050628769 LRGDP0 -0.024825369 0.038398138 RRA -0.054767167 0.34136609 PREMIUM -7.1061187E-005 3.3475029E-005 Observations 97.00000 Degrees of Freedom 90.00000 Sum of Squared Errors 0.03626 Residual Variance 0.00040 R-squared 0.29641 _____________________________________________ Regime 2: RRA > 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.30403012 0.058022133 LS 0.022302199 0.0046377621 LPRIMARY 0.049108721 0.010214662 LPOP 0.0021269636 0.018358440 LRGDP0 -0.036674165 0.0057669973 RRA 0.0040047783 0.0018252154 PREMIUM -0.00057967521 7.1666488E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.095546518 0.43235329 LS -0.0061865593 0.037559491 LPRIMARY -0.029625500 0.10466284 LPOP -0.10275411 0.039410428 LRGDP0 -0.062547272 -0.011591286 RRA -0.0033382504 0.013393825 PREMIUM -0.00083455584 0.0084876888 Observations 41.00000 Degrees of Freedom 34.00000 Sum of Squared Errors 0.00261 Residual Variance 0.00008 R-squared 0.80522 __________________________________________________ --------------------- Sensitivity specification (6d) --------------------------- -------- Baseline L_M growth regression (specification 6d in distortthr09.out) plus RRA and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 8.19420 LM-test for no threshold 31.24582 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01180 LM-test for no threshold 25.49509 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 8.7714080 2.6703447 LM0 0.49756897 0.35439644 LS 0.031555904 0.18181997 LPRIMARY 0.78863933 0.45478205 LPOP 1.8226488 1.1454119 LRGDP0 -0.58004293 0.21057491 RRA 0.50144345 0.30823859 PREMIUM -0.0011478653 0.00069027617 Observations 118.00000 Degrees of Freedom 110.00000 Sum of Squared Errors 105.40529 Residual Variance 0.95823 R-squared 0.16008 Heteroskedasticity Test (P-Value) 0.48350 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 60.77223 Residual Variance 0.59581 Joint R-Squared 0.51574 Heteroskedasticity Test (p-value) 0.02057 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.0385866 3.4214504 LM0 0.057981222 0.33326556 LS -0.077604002 0.18322114 LPRIMARY 0.21581842 0.41793283 LPOP 0.098119382 1.1668238 LRGDP0 0.40813318 0.24672305 RRA 1.7801691 0.42021649 PREMIUM -0.00035746003 0.00042811428 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -7.7446294 7.2064978 LM0 -0.64933075 0.71118173 LS -0.43671744 0.29326580 LPRIMARY -0.60332992 1.0885549 LPOP -2.1888552 2.7801893 LRGDP0 -0.15244099 0.89171037 RRA 0.95654480 2.6618873 PREMIUM -0.0012917740 0.00050989695 Observations 67.00000 Degrees of Freedom 59.00000 Sum of Squared Errors 56.14128 Residual Variance 0.95155 R-squared 0.40565 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.778321 3.2880015 LM0 -4.3718721 0.98451989 LS 0.66100968 0.21347566 LPRIMARY -2.4286714 1.8513038 LPOP 0.20175908 0.54273026 LRGDP0 -0.036610564 0.24214593 RRA -0.034016587 0.067465251 PREMIUM -0.0024205557 0.0025806317 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 9.2803011 25.366762 LM0 -6.3015311 -2.0976957 LS 0.19995394 1.0794220 LPRIMARY -6.0572268 3.1159194 LPOP -1.1009471 1.2655104 LRGDP0 -0.56278369 0.43799545 RRA -0.16624848 0.11207509 PREMIUM -0.0076385495 0.0068035148 Observations 51.00000 Degrees of Freedom 43.00000 Sum of Squared Errors 4.63096 Residual Variance 0.10770 R-squared 0.77198 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 8.7714080 2.6703447 LM0 0.49756897 0.35439644 LS 0.031555904 0.18181997 LPRIMARY 0.78863933 0.45478205 LPOP 1.8226488 1.1454119 LRGDP0 -0.58004293 0.21057491 RRA 0.50144345 0.30823859 PREMIUM -0.0011478653 0.00069027617 Observations 118.00000 Degrees of Freedom 110.00000 Sum of Squared Errors 105.40529 Residual Variance 0.95823 R-squared 0.16008 Heteroskedasticity Test (P-Value) 0.48350 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.01180 95% Confidence Interval: [ 0.011800 , 0.110800 ] Sum of Squared Errors 72.65680 Residual Variance 0.71232 Joint R-Squared 0.42104 Heteroskedasticity Test (p-value) 0.42373 _____________________________________________ Regime 1: RRA <= 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 5.3399191 4.1201579 LM0 0.14061164 0.32810033 LS -0.11010017 0.16009424 LPRIMARY 0.71734485 0.32092351 LPOP 0.99634723 1.7240335 LRGDP0 -0.18433140 0.20522449 RRA 1.1322367 0.59184171 PREMIUM -0.00086959173 0.00049475942 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.7355903 15.379326 LM0 -0.53663832 0.82766585 LS -0.42388488 0.26321274 LPRIMARY 0.088334779 1.4581332 LPOP -2.3925218 5.2226020 LRGDP0 -0.69446336 0.22669935 RRA -0.36404689 2.2922465 PREMIUM -0.0019872328 0.00011358537 Observations 72.00000 Degrees of Freedom 64.00000 Sum of Squared Errors 53.42081 Residual Variance 0.83470 R-squared 0.15073 _____________________________________________ Regime 2: RRA > 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 22.487731 5.4857122 LM0 -4.1302285 1.5823394 LS 1.5469580 0.66114293 LPRIMARY 1.6770210 1.9264328 LPOP 1.5320998 1.0369511 LRGDP0 -0.43258630 0.55210557 RRA 0.0081373982 0.21975753 PREMIUM -0.0049142166 0.0075578742 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 11.174446 33.472008 LM0 -7.3210937 -0.48869487 LS 0.20680684 3.1776069 LPRIMARY -2.2337935 5.6174130 LPOP -1.3816079 3.7226332 LRGDP0 -1.9859541 0.64954061 RRA -0.43223143 0.47875732 PREMIUM -0.021056616 0.0099402614 Observations 46.00000 Degrees of Freedom 38.00000 Sum of Squared Errors 19.23599 Residual Variance 0.50621 R-squared 0.69229 __________________________________________________ --------------------- Senstivity specification (7d) ---------------------------- ------- Baseline output growth regression (specification 5d in distortthr09.out) plus RRP and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 22.55616 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.35950 LM-test for no threshold 20.78083 Bootstrap P-Value 0.00400 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.024376764 0.042396653 LS 0.011713334 0.0042631896 LPRIMARY 0.016584685 0.0076736509 LPOP -0.050949588 0.018776697 LRGDP0 -0.014273097 0.0039158974 RRA 0.0069869010 0.0040521769 PI -2.4426167E-005 1.8741588E-005 Observations 140.00000 Degrees of Freedom 133.00000 Sum of Squared Errors 0.05487 Residual Variance 0.00041 R-squared 0.26736 Heteroskedasticity Test (P-Value) 0.00660 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.19676 95% Confidence Interval: [ 6.710644 , 8.837739 ] Sum of Squared Errors 0.04458 Residual Variance 0.00035 Joint R-Squared 0.40479 Heteroskedasticity Test (p-value) 0.00916 _____________________________________________ Regime 1: LRGDP0 <= 7.196761 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.0023983187 0.19996572 LS 0.0062671401 0.0059007545 LPRIMARY -0.0018093769 0.013259378 LPOP -0.082602457 0.059728623 LRGDP0 -0.029604558 0.023303247 RRA 0.019723995 0.013107174 PI -4.0032867E-005 2.0013050E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.48983013 1.4638237 LS -0.025015164 0.019548213 LPRIMARY -0.058164127 0.043568446 LPOP -0.50095212 0.43072637 LRGDP0 -0.21430157 0.087261007 RRA -0.066095735 0.18292553 PI -0.00012789872 3.3291536E-005 Observations 34.00000 Degrees of Freedom 27.00000 Sum of Squared Errors 0.01578 Residual Variance 0.00058 R-squared 0.24032 _____________________________________________ Regime 2: LRGDP0 > 7.196761 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.031911740 0.039156133 LS 0.019836304 0.0051751984 LPRIMARY 0.029839062 0.0087695029 LPOP -0.054646757 0.017934002 LRGDP0 -0.024967990 0.0040286685 RRA 0.0059771985 0.0035588014 PI 5.2067350E-005 8.6899061E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.087478916 0.28348029 LS 0.0026026493 0.032965582 LPRIMARY -0.058512417 0.094032749 LPOP -0.093627247 0.011767989 LRGDP0 -0.043322441 -0.0086987098 RRA -0.0031880944 0.015891741 PI -0.00031444720 0.00060618971 Observations 106.00000 Degrees of Freedom 99.00000 Sum of Squared Errors 0.02880 Residual Variance 0.00029 R-squared 0.40421 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.024376764 0.042396653 LS 0.011713334 0.0042631896 LPRIMARY 0.016584685 0.0076736509 LPOP -0.050949588 0.018776697 LRGDP0 -0.014273097 0.0039158974 RRA 0.0069869010 0.0040521769 PI -2.4426167E-005 1.8741588E-005 Observations 140.00000 Degrees of Freedom 133.00000 Sum of Squared Errors 0.05487 Residual Variance 0.00041 R-squared 0.26736 Heteroskedasticity Test (P-Value) 0.00660 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.17840 95% Confidence Interval: [ -0.62320 , 0.359500 ] Sum of Squared Errors 0.04356 Residual Variance 0.00035 Joint R-Squared 0.41848 Heteroskedasticity Test (p-value) 0.00305 _____________________________________________ Regime 1: RRA <= 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.049632472 0.049029741 LS 0.0097741926 0.0040991721 LPRIMARY 0.012503115 0.0081103040 LPOP -0.071297506 0.017674537 LRGDP0 -0.016482699 0.0035657603 RRA 0.035159258 0.011721003 PI -1.6746484E-005 1.9231933E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.3109278 0.29340691 LS -0.0032757307 0.031356448 LPRIMARY -0.059003104 0.030319980 LPOP -0.50189689 0.046685088 LRGDP0 -0.023846614 0.036931872 RRA 0.0043197824 0.39170851 PI -6.4888777E-005 2.2118876E-005 Observations 97.00000 Degrees of Freedom 90.00000 Sum of Squared Errors 0.03960 Residual Variance 0.00044 R-squared 0.30389 _____________________________________________ Regime 2: RRA > 0.178400 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.24044004 0.054892336 LS 0.022050197 0.0053684459 LPRIMARY 0.0044481434 0.010772799 LPOP 0.022143920 0.022858139 LRGDP0 -0.021029165 0.0057571473 RRA 0.0032813449 0.0023110689 PI -0.00030901167 5.4289264E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.10247578 0.34802902 LS 0.0069920350 0.041121896 LPRIMARY -0.019888139 0.042693144 LPOP -0.099751383 0.066945873 LRGDP0 -0.039027835 -0.0091530646 RRA -0.0036501581 0.012256412 PI -0.00042093935 0.00012099739 Observations 43.00000 Degrees of Freedom 36.00000 Sum of Squared Errors 0.00395 Residual Variance 0.00011 R-squared 0.73374 __________________________________________________ --------------------- Senstivity specification (8d) ---------------------------- ------- Baseline L_M growth regression (specification 6d in distortthr09.out) plus RRA and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 30.04106 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using RRA Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate -0.05990 LM-test for no threshold 25.55774 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.8683518 2.3894104 LM0 0.55396160 0.36277477 LS 0.19210616 0.21726943 LPRIMARY 0.69767950 0.48793343 LPOP 1.2506345 1.0078519 LRGDP0 -0.63016795 0.21179832 RRA 0.33430006 0.25686755 PI 0.00052353879 0.0032611861 Observations 119.00000 Degrees of Freedom 111.00000 Sum of Squared Errors 111.36755 Residual Variance 1.00331 R-squared 0.12964 Heteroskedasticity Test (P-Value) 0.35749 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.474243 ] Sum of Squared Errors 57.92617 Residual Variance 0.56239 Joint R-Squared 0.54729 Heteroskedasticity Test (p-value) 0.01778 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -4.8221664 3.6737414 LM0 0.036724931 0.31104068 LS 0.24103251 0.22875980 LPRIMARY 0.038772396 0.43699634 LPOP -0.33288752 0.93184309 LRGDP0 0.59740624 0.26108034 RRA 1.5418892 0.39205813 PI 0.0048641888 0.0032052546 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -12.022700 2.3783668 LM0 -0.57291479 0.64636466 LS -0.20733670 0.68940172 LPRIMARY -0.81774042 0.89528522 LPOP -2.1593000 1.4935249 LRGDP0 0.085688773 1.1091237 RRA 0.77345524 2.3103231 PI -0.0014181102 0.011146488 Observations 67.00000 Degrees of Freedom 59.00000 Sum of Squared Errors 53.21429 Residual Variance 0.90194 R-squared 0.43736 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.246510 3.2636078 LM0 -4.2994755 0.93926329 LS 0.58724621 0.19971996 LPRIMARY -3.0214481 2.0206250 LPOP 0.056979302 0.53772689 LRGDP0 0.11219625 0.22644434 RRA -0.048938124 0.069023360 PI -0.0046112938 0.0025284568 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 10.849839 23.643181 LM0 -6.1404316 -2.4585195 LS 0.19579509 0.97869733 LPRIMARY -6.9818730 0.93897678 LPOP -0.99696541 1.1109240 LRGDP0 -0.33163465 0.55602715 RRA -0.18422391 0.086347661 PI -0.0095670692 0.00034448156 Observations 52.00000 Degrees of Freedom 44.00000 Sum of Squared Errors 4.71189 Residual Variance 0.10709 R-squared 0.77276 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.8683518 2.3894104 LM0 0.55396160 0.36277477 LS 0.19210616 0.21726943 LPRIMARY 0.69767950 0.48793343 LPOP 1.2506345 1.0078519 LRGDP0 -0.63016795 0.21179832 RRA 0.33430006 0.25686755 PI 0.00052353879 0.0032611861 Observations 119.00000 Degrees of Freedom 111.00000 Sum of Squared Errors 111.36755 Residual Variance 1.00331 R-squared 0.12964 Heteroskedasticity Test (P-Value) 0.35749 __________________________________________________ Threshold Estimation Threshold Variable = RRA Threshold Estimate 0.01180 95% Confidence Interval: [ 0.011800 , 0.110800 ] Sum of Squared Errors 72.41531 Residual Variance 0.70306 Joint R-Squared 0.43406 Heteroskedasticity Test (p-value) 0.46616 _____________________________________________ Regime 1: RRA <= 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 6.4333169 4.1333699 LM0 0.10567675 0.31685427 LS -0.20716323 0.18048277 LPRIMARY 0.74991808 0.32666060 LPOP 1.3373080 1.7502567 LRGDP0 -0.12633350 0.19427188 RRA 1.2331991 0.56455400 PI -0.0038168206 0.0019205689 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.6680881 15.639070 LM0 -0.54604309 0.74288553 LS -0.56090945 0.18087559 LPRIMARY 0.10966330 1.5103557 LPOP -2.0931951 5.2617691 LRGDP0 -0.66649724 0.26312640 RRA -0.051052900 2.3397250 PI -0.0082978168 -5.2505629E-005 Observations 71.00000 Degrees of Freedom 63.00000 Sum of Squared Errors 52.79804 Residual Variance 0.83806 R-squared 0.15979 _____________________________________________ Regime 2: RRA > 0.011800 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.775578 5.3141949 LM0 -4.0736378 1.8309952 LS 1.7772128 0.63533448 LPRIMARY 1.5231637 1.6567569 LPOP 0.69078570 0.75340065 LRGDP0 -0.41842013 0.55819516 RRA -0.15964939 0.19308764 PI 0.0030202675 0.0052327670 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 8.1298514 30.200817 LM0 -8.3010751 0.025188563 LS 0.47809434 3.4047492 LPRIMARY -1.9765994 4.9223124 LPOP -1.6797535 2.7017480 LRGDP0 -1.8891680 0.83332392 RRA -0.56527354 0.31964722 PI -0.0082614528 0.013276491 Observations 48.00000 Degrees of Freedom 40.00000 Sum of Squared Errors 19.61727 Residual Variance 0.49043 R-squared 0.69818 __________________________________________________ E) Border market support: 1960 -- 2000 --------------------- Senstivity specification (1e) ---------------------------- ------- Baseline output growth regression (specification 5e in distortthr09.out) plus BMS as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.33200 LM-test for no threshold 35.42057 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.47050 LM-test for no threshold 25.62473 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.062011022 0.036775271 LS 0.016249018 0.0032928512 LPRIMARY 0.010188978 0.0046872225 LPOP -0.052940177 0.016865188 LRGDP0 -0.012577373 0.0029692306 BMS 0.0053307345 0.0035630197 Observations 211.00000 Degrees of Freedom 205.00000 Sum of Squared Errors 0.08029 Residual Variance 0.00039 R-squared 0.26884 Heteroskedasticity Test (P-Value) 0.00051 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 7.228315 , 8.900249 ] Sum of Squared Errors 0.06611 Residual Variance 0.00033 Joint R-Squared 0.39795 Heteroskedasticity Test (p-value) 0.00042 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.22307294 0.062523657 LS 0.015166821 0.0033497647 LPRIMARY 0.0038129258 0.0052095160 LPOP -0.097362560 0.019131730 LRGDP0 -0.0059003626 0.0055159064 BMS 0.029121574 0.0083916159 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.39402565 0.15346816 LS 0.0043517951 0.021878519 LPRIMARY -0.011426345 0.015329147 LPOP -0.18605462 -0.013016203 LRGDP0 -0.048275497 0.0051619420 BMS 0.0052993683 0.062837924 Observations 120.00000 Degrees of Freedom 114.00000 Sum of Squared Errors 0.05069 Residual Variance 0.00044 R-squared 0.42972 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.087670494 0.051745309 LS 0.0091648405 0.0056451337 LPRIMARY 0.0014872765 0.024982940 LPOP -0.016151769 0.018067723 LRGDP0 -0.015127552 0.0030553778 BMS 0.0051067693 0.0031259584 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.064501266 0.23405271 LS -0.010831617 0.027128304 LPRIMARY -0.047479286 0.069410031 LPOP -0.083767987 0.020551897 LRGDP0 -0.026457204 -0.0082481554 BMS -0.0016703813 0.012896191 Observations 91.00000 Degrees of Freedom 85.00000 Sum of Squared Errors 0.01543 Residual Variance 0.00018 R-squared 0.18360 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.062011022 0.036775271 LS 0.016249018 0.0032928512 LPRIMARY 0.010188978 0.0046872225 LPOP -0.052940177 0.016865188 LRGDP0 -0.012577373 0.0029692306 BMS 0.0053307345 0.0035630197 Observations 211.00000 Degrees of Freedom 205.00000 Sum of Squared Errors 0.08029 Residual Variance 0.00039 R-squared 0.26884 Heteroskedasticity Test (P-Value) 0.00051 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.094100 , 0.595300 ] Sum of Squared Errors 0.06594 Residual Variance 0.00033 Joint R-Squared 0.39950 Heteroskedasticity Test (p-value) 0.00035 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.11033556 0.055770011 LS 0.014635233 0.0032435786 LPRIMARY 0.0062889762 0.0046150902 LPOP -0.065941016 0.023114667 LRGDP0 -0.010090186 0.0031682344 BMS 0.030724265 0.012564950 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.22422911 0.0035664968 LS 0.0082778194 0.021594218 LPRIMARY -0.0027566005 0.016813625 LPOP -0.11664658 -0.020636269 LRGDP0 -0.017964185 -0.0037362935 BMS 0.0054347425 0.061185726 Observations 136.00000 Degrees of Freedom 130.00000 Sum of Squared Errors 0.05469 Residual Variance 0.00042 R-squared 0.24522 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.14807803 0.058795154 LS 0.022165554 0.0041683971 LPRIMARY 0.034942306 0.0089192381 LPOP -0.023592984 0.022035336 LRGDP0 -0.027730677 0.0034468612 BMS 0.0021186831 0.0023642145 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.039888222 0.44537618 LS -0.0035025347 0.034816973 LPRIMARY -0.024334671 0.18391483 LPOP -0.091062264 0.059086095 LRGDP0 -0.040416993 -0.014607054 BMS -0.0047205620 0.010352491 Observations 75.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 0.01125 Residual Variance 0.00016 R-squared 0.57001 __________________________________________________ --------------------- Sensitivity specification (2e) --------------------------- -------- Baseline L_M growth regression (specification 6e in distortthr09.out) plus BMS as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 42.71976 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01300 LM-test for no threshold 28.65068 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.9737630 1.7783786 LM0 0.53386814 0.25401638 LS 0.21700781 0.12213577 LPRIMARY 0.43093475 0.32266691 LPOP 1.2784656 0.80848945 LRGDP0 -0.52314219 0.15591458 BMS 0.24561885 0.17823678 Observations 186.00000 Degrees of Freedom 179.00000 Sum of Squared Errors 124.48807 Residual Variance 0.69546 R-squared 0.14933 Heteroskedasticity Test (P-Value) 0.28022 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.48322 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 76.97795 Residual Variance 0.44755 Joint R-Squared 0.47399 Heteroskedasticity Test (p-value) 0.06265 _____________________________________________ Regime 1: LRGDP0 <= 8.483222 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.63639242 2.2842460 LM0 0.16956331 0.24119645 LS 0.026365052 0.098423216 LPRIMARY 0.12476347 0.29414705 LPOP 1.0501860 0.90927159 LRGDP0 0.36578082 0.17548746 BMS 1.8084918 0.33803115 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -4.6707500 5.5517593 LM0 -0.31409677 0.65931516 LS -0.17191160 0.24184848 LPRIMARY -0.48141466 0.70560598 LPOP -0.86230045 2.9900420 LRGDP0 0.021825407 0.76983702 BMS 1.1120666 2.4710328 Observations 107.00000 Degrees of Freedom 100.00000 Sum of Squared Errors 71.03097 Residual Variance 0.71031 R-squared 0.34521 _____________________________________________ Regime 2: LRGDP0 > 8.483222 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 16.179464 2.6361156 LM0 -3.8978980 0.77720387 LS 0.58822777 0.14941639 LPRIMARY -3.0695317 2.1097648 LPOP -0.091488494 0.41581460 LRGDP0 -0.045547520 0.13947781 BMS -0.078220360 0.053598675 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.9627173 21.346251 LM0 -5.4212176 0.077299654 LS 0.22172378 0.99634622 LPRIMARY -7.3663514 3.8756382 LPOP -1.0900002 1.6873937 LRGDP0 -0.67410226 0.22782899 BMS -0.20096630 0.10279682 Observations 79.00000 Degrees of Freedom 72.00000 Sum of Squared Errors 5.94698 Residual Variance 0.08260 R-squared 0.75574 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.9737630 1.7783786 LM0 0.53386814 0.25401638 LS 0.21700781 0.12213577 LPRIMARY 0.43093475 0.32266691 LPOP 1.2784656 0.80848945 LRGDP0 -0.52314219 0.15591458 BMS 0.24561885 0.17823678 Observations 186.00000 Degrees of Freedom 179.00000 Sum of Squared Errors 124.48807 Residual Variance 0.69546 R-squared 0.14933 Heteroskedasticity Test (P-Value) 0.28022 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.080100 , 0.102300 ] Sum of Squared Errors 91.29113 Residual Variance 0.53076 Joint R-Squared 0.37618 Heteroskedasticity Test (p-value) 0.15421 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 9.0347240 2.2364616 LM0 0.46962887 0.26610276 LS 0.050629461 0.10004984 LPRIMARY 0.10853164 0.31659202 LPOP 3.3261127 0.93765428 LRGDP0 -0.14903434 0.13323426 BMS 0.80186399 0.42369260 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 4.6512592 13.527387 LM0 -0.051932546 1.0033315 LS -0.14546822 0.25141456 LPRIMARY -0.51198872 0.75612603 LPOP 1.4757366 5.1639150 LRGDP0 -0.43695705 0.11210481 BMS -0.028573499 1.7348551 Observations 119.00000 Degrees of Freedom 112.00000 Sum of Squared Errors 72.64783 Residual Variance 0.64864 R-squared 0.14407 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 16.652260 4.1318446 LM0 -4.1542993 1.0723864 LS 0.80236724 0.33087577 LPRIMARY 3.6307875 0.68184343 LPOP -0.79555658 0.62624045 LRGDP0 -0.26723398 0.32182457 BMS 0.13247091 0.14205000 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 8.2226891 26.877256 LM0 -6.3650941 -2.0524219 LS 0.085635899 1.4508837 LPRIMARY 2.2121839 5.4757789 LPOP -2.0894291 0.61547205 LRGDP0 -0.89801014 0.36354218 BMS -0.14594709 0.42933774 Observations 67.00000 Degrees of Freedom 60.00000 Sum of Squared Errors 18.64330 Residual Variance 0.31072 R-squared 0.69074 __________________________________________________ --------------------- Senstivity specification (5e) ---------------------------- ------- Baseline output growth regression (specification 5e in distortthr09.out) plus RRP and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.22710 LM-test for no threshold 33.71026 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.47050 LM-test for no threshold 25.46055 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.080441451 0.030606994 LS 0.013943012 0.0032594602 LPRIMARY 0.0093159261 0.0045498820 LPOP -0.062578874 0.013875525 LRGDP0 -0.012655812 0.0027689007 BMS 0.0035102879 0.0033978466 PREMIUM -3.2200628E-005 1.2306322E-005 Observations 207.00000 Degrees of Freedom 200.00000 Sum of Squared Errors 0.07181 Residual Variance 0.00036 R-squared 0.29621 Heteroskedasticity Test (P-Value) 0.00500 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.72590 95% Confidence Interval: [ 7.718818 , 8.474243 ] Sum of Squared Errors 0.05832 Residual Variance 0.00030 Joint R-Squared 0.42844 Heteroskedasticity Test (p-value) 0.03097 _____________________________________________ Regime 1: LRGDP0 <= 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.25084395 0.10152107 LS 0.012757564 0.0036747640 LPRIMARY 0.0048210672 0.0051319971 LPOP -0.12020386 0.029203185 LRGDP0 -0.0095159343 0.0074630526 BMS 0.027316722 0.013619553 PREMIUM -1.7540674E-005 8.6943191E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.44982524 -0.051862656 LS 0.0055550265 0.019960102 LPRIMARY -0.0052376471 0.014879781 LPOP -0.17744211 -0.062965622 LRGDP0 -0.024143517 0.0051116487 BMS 0.00062239847 0.054011045 PREMIUM -3.4581539E-005 -4.9980852E-007 Observations 88.00000 Degrees of Freedom 81.00000 Sum of Squared Errors 0.03307 Residual Variance 0.00041 R-squared 0.32875 _____________________________________________ Regime 2: LRGDP0 > 7.725903 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.045606692 0.036845179 LS 0.015795764 0.0043700009 LPRIMARY -0.032331386 0.024927857 LPOP -0.042589821 0.013332363 LRGDP0 -0.020605072 0.0034694858 BMS 0.0023556223 0.0027525411 PREMIUM -0.00034842190 8.6876850E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.026609858 0.11782324 LS 0.0072305624 0.024360966 LPRIMARY -0.081189985 0.016527214 LPOP -0.068721252 -0.016458390 LRGDP0 -0.027405264 -0.013804879 BMS -0.0030393582 0.0077506027 PREMIUM -0.00051870052 -0.00017814327 Observations 119.00000 Degrees of Freedom 112.00000 Sum of Squared Errors 0.02525 Residual Variance 0.00023 R-squared 0.39260 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.080441451 0.030606994 LS 0.013943012 0.0032594602 LPRIMARY 0.0093159261 0.0045498820 LPOP -0.062578874 0.013875525 LRGDP0 -0.012655812 0.0027689007 BMS 0.0035102879 0.0033978466 PREMIUM -3.2200628E-005 1.2306322E-005 Observations 207.00000 Degrees of Freedom 200.00000 Sum of Squared Errors 0.07181 Residual Variance 0.00036 R-squared 0.29621 Heteroskedasticity Test (P-Value) 0.00500 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ -0.43690 , 0.640900 ] Sum of Squared Errors 0.06145 Residual Variance 0.00032 Joint R-Squared 0.39779 Heteroskedasticity Test (p-value) 0.00389 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.13703413 0.053762855 LS 0.012917640 0.0032577537 LPRIMARY 0.0055224695 0.0044930584 LPOP -0.076726545 0.022763838 LRGDP0 -0.0098157977 0.0031024124 BMS 0.020396258 0.012663114 PREMIUM -2.2619515E-005 1.0221786E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.25606864 0.024733220 LS 0.0062389164 0.020712553 LPRIMARY -0.0039153401 0.018569984 LPOP -0.12894031 -0.017369311 LRGDP0 -0.022222661 -0.0010753479 BMS -0.0050873891 0.062463581 PREMIUM -4.9538124E-005 2.6271162E-006 Observations 135.00000 Degrees of Freedom 128.00000 Sum of Squared Errors 0.05128 Residual Variance 0.00040 R-squared 0.26200 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.11078646 0.061055563 LS 0.020359962 0.0042324479 LPRIMARY 0.033417233 0.010312045 LPOP -0.034620111 0.020622414 LRGDP0 -0.026600853 0.0040663916 BMS 0.0025923423 0.0025106961 PREMIUM 6.8885032E-005 0.00022397482 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.084846297 0.56085802 LS -0.010757280 0.032058642 LPRIMARY -0.048939065 0.17667721 LPOP -0.10640446 0.099695738 LRGDP0 -0.045070877 -0.012775178 BMS -0.0054911643 0.011757729 PREMIUM -0.0013690416 0.010594873 Observations 72.00000 Degrees of Freedom 65.00000 Sum of Squared Errors 0.01017 Residual Variance 0.00016 R-squared 0.54647 __________________________________________________ --------------------- Sensitivity specification (6e) --------------------------- -------- Baseline L_M growth regression (specification 6e in distortthr09.out) plus BMS and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 43.54285 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01300 LM-test for no threshold 30.02163 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.8363881 1.8542345 LM0 0.50165717 0.25228562 LS 0.17051500 0.12769120 LPRIMARY 0.44646982 0.32050841 LPOP 1.6335725 0.84113969 LRGDP0 -0.47863121 0.15653522 BMS 0.28565878 0.18847674 PREMIUM -0.00075468223 0.00049183939 Observations 183.00000 Degrees of Freedom 175.00000 Sum of Squared Errors 120.50353 Residual Variance 0.68859 R-squared 0.15920 Heteroskedasticity Test (P-Value) 0.38948 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 76.22176 Residual Variance 0.45642 Joint R-Squared 0.46817 Heteroskedasticity Test (p-value) 0.07521 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 1.5957069 2.6783460 LM0 0.16034764 0.24291588 LS 0.036639198 0.10467181 LPRIMARY 0.10988763 0.29532191 LPOP 1.6077505 1.0746247 LRGDP0 0.43228833 0.17532629 BMS 1.8170841 0.35445909 PREMIUM 6.9443338E-005 0.00033220841 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -3.7547545 7.1840674 LM0 -0.32995069 0.63646275 LS -0.17667842 0.25749237 LPRIMARY -0.46894332 0.70716023 LPOP -0.69027541 3.7317435 LRGDP0 0.043868662 0.77592786 BMS 1.1115668 2.5250815 PREMIUM -0.00059178008 0.00073172890 Observations 106.00000 Degrees of Freedom 98.00000 Sum of Squared Errors 71.17258 Residual Variance 0.72625 R-squared 0.35267 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 14.508576 2.4006179 LM0 -3.5518054 0.74586022 LS 0.59017681 0.15153701 LPRIMARY -0.75756455 1.4448307 LPOP -0.26238201 0.33187114 LRGDP0 -0.084434107 0.14096567 BMS -0.047142934 0.046192507 PREMIUM 0.00061898884 0.0040828950 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 9.2230086 21.600489 LM0 -5.4315487 -2.0769484 LS 0.29316426 0.98879587 LPRIMARY -7.0235826 2.0743036 LPOP -0.93016732 0.83243959 LRGDP0 -0.36072682 0.24670721 BMS -0.15403490 0.043394381 PREMIUM -0.0078070798 0.0087050787 Observations 77.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 5.04918 Residual Variance 0.07318 R-squared 0.73413 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.8363881 1.8542345 LM0 0.50165717 0.25228562 LS 0.17051500 0.12769120 LPRIMARY 0.44646982 0.32050841 LPOP 1.6335725 0.84113969 LRGDP0 -0.47863121 0.15653522 BMS 0.28565878 0.18847674 PREMIUM -0.00075468223 0.00049183939 Observations 183.00000 Degrees of Freedom 175.00000 Sum of Squared Errors 120.50353 Residual Variance 0.68859 R-squared 0.15920 Heteroskedasticity Test (P-Value) 0.38948 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.080100 , 0.351300 ] Sum of Squared Errors 90.26781 Residual Variance 0.54053 Joint R-Squared 0.37016 Heteroskedasticity Test (p-value) 0.21639 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 8.9785114 2.3145519 LM0 0.48585768 0.26667236 LS 0.024537137 0.10635326 LPRIMARY 0.11175805 0.31678721 LPOP 3.2661584 0.97024432 LRGDP0 -0.15996864 0.13415211 BMS 0.69272214 0.46326725 PREMIUM -0.00039632120 0.00046755554 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 1.4579706 13.647435 LM0 -0.12117817 1.0220982 LS -0.18391525 0.29009320 LPRIMARY -0.50914489 0.93571756 LPOP -0.49626604 5.1678373 LRGDP0 -0.57354020 0.10296949 BMS -0.21528167 1.7074377 PREMIUM -0.0013160898 0.00054837808 Observations 118.00000 Degrees of Freedom 110.00000 Sum of Squared Errors 72.34660 Residual Variance 0.65770 R-squared 0.14340 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.779978 4.3512340 LM0 -4.0138213 1.0640427 LS 0.91779665 0.34911764 LPRIMARY 3.1049351 0.68195615 LPOP -0.34146372 0.74101158 LRGDP0 -0.35269200 0.39319843 BMS 0.14474818 0.14167489 PREMIUM -0.014173567 0.011463130 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 8.8381539 38.496149 LM0 -7.5490173 -1.9267228 LS -0.17283667 1.6020672 LPRIMARY -1.4205024 10.393402 LPOP -2.9803417 2.3302891 LRGDP0 -1.3593756 0.41972675 BMS -0.24390036 0.48660068 PREMIUM -0.11023454 0.010827858 Observations 65.00000 Degrees of Freedom 57.00000 Sum of Squared Errors 17.92121 Residual Variance 0.31441 R-squared 0.68916 __________________________________________________ --------------------- Senstivity specification (7e) ---------------------------- ------- Baseline output growth regression (specification 5e in distortthr09.out) plus RRP and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 35.79537 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.29590 LM-test for no threshold 26.42254 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.059105543 0.036436668 LS 0.015423883 0.0034706806 LPRIMARY 0.0093037719 0.0046975125 LPOP -0.052605094 0.016901602 LRGDP0 -0.012394351 0.0030391442 BMS 0.0056219553 0.0035521921 PI -1.5371536E-005 2.0660677E-005 Observations 209.00000 Degrees of Freedom 202.00000 Sum of Squared Errors 0.07949 Residual Variance 0.00039 R-squared 0.27190 Heteroskedasticity Test (P-Value) 0.00106 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 6.463653 , 8.900249 ] Sum of Squared Errors 0.06578 Residual Variance 0.00034 Joint R-Squared 0.39744 Heteroskedasticity Test (p-value) 0.00019 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.23033487 0.062175700 LS 0.016124398 0.0033989392 LPRIMARY 0.0032415917 0.0052578120 LPOP -0.098561039 0.019025524 LRGDP0 -0.0057738284 0.0054958786 BMS 0.028041874 0.0084622906 PI 9.5210384E-006 2.8955018E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.39713796 0.19314215 LS 0.0025879223 0.022897073 LPRIMARY -0.012471925 0.015456731 LPOP -0.18900175 0.00057602166 LRGDP0 -0.049022229 0.0052965513 BMS 0.0022062162 0.062166561 PI -5.7153729E-005 6.6535068E-005 Observations 118.00000 Degrees of Freedom 111.00000 Sum of Squared Errors 0.05036 Residual Variance 0.00045 R-squared 0.43042 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.085421732 0.058026201 LS 0.0087898996 0.0059642695 LPRIMARY 0.0014562796 0.024852262 LPOP -0.016062052 0.018047934 LRGDP0 -0.014593963 0.0047669951 BMS 0.0052888669 0.0029649541 PI -1.6706671E-005 0.00010180857 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.075598181 0.25572478 LS -0.011126469 0.031454541 LPRIMARY -0.047254154 0.069174354 LPOP -0.084018021 0.019311899 LRGDP0 -0.029595358 -0.0040593359 BMS -0.0016823292 0.013027841 PI -0.00023172266 0.00027335088 Observations 91.00000 Degrees of Freedom 84.00000 Sum of Squared Errors 0.01542 Residual Variance 0.00018 R-squared 0.18381 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.059105543 0.036436668 LS 0.015423883 0.0034706806 LPRIMARY 0.0093037719 0.0046975125 LPOP -0.052605094 0.016901602 LRGDP0 -0.012394351 0.0030391442 BMS 0.0056219553 0.0035521921 PI -1.5371536E-005 2.0660677E-005 Observations 209.00000 Degrees of Freedom 202.00000 Sum of Squared Errors 0.07949 Residual Variance 0.00039 R-squared 0.27190 Heteroskedasticity Test (P-Value) 0.00106 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.054500 , 0.595300 ] Sum of Squared Errors 0.06570 Residual Variance 0.00034 Joint R-Squared 0.39817 Heteroskedasticity Test (p-value) 0.00047 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.10898697 0.055722461 LS 0.014596199 0.0032703850 LPRIMARY 0.0058046612 0.0046855393 LPOP -0.065498405 0.023216694 LRGDP0 -0.010079245 0.0032241272 BMS 0.029829746 0.012803677 PI -3.1360440E-006 2.3462284E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.22268752 0.0051625388 LS 0.0079410957 0.021434372 LPRIMARY -0.0033789957 0.016290368 LPOP -0.11650451 -0.019993685 LRGDP0 -0.017999088 -0.0036209831 BMS 0.0044069132 0.060763182 PI -5.2618620E-005 4.2891945E-005 Observations 134.00000 Degrees of Freedom 127.00000 Sum of Squared Errors 0.05446 Residual Variance 0.00043 R-squared 0.24517 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.14942730 0.058275760 LS 0.023001367 0.0060121887 LPRIMARY 0.035716749 0.0089699410 LPOP -0.023560188 0.021739753 LRGDP0 -0.028302982 0.0042580947 BMS 0.0018431282 0.0024732754 PI 1.8860144E-005 7.5937780E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.034230655 0.46210834 LS -0.0032598083 0.040862055 LPRIMARY -0.023052801 0.15879458 LPOP -0.089406263 0.066627479 LRGDP0 -0.048030230 -0.0088084263 BMS -0.0041254272 0.0096646488 PI -0.00027673637 0.00027292449 Observations 75.00000 Degrees of Freedom 68.00000 Sum of Squared Errors 0.01124 Residual Variance 0.00017 R-squared 0.57027 __________________________________________________ --------------------- Senstivity specification (8e) ---------------------------- ------- Baseline L_M growth regression (specification 6e in distortthr09.out) plus BMS and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 39.26104 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01300 LM-test for no threshold 33.76922 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.8198033 1.7748941 LM0 0.55274277 0.26381375 LS 0.26578408 0.17516247 LPRIMARY 0.41434997 0.32857407 LPOP 1.2471748 0.79609285 LRGDP0 -0.54771397 0.16746661 NRP 0.22160566 0.18797398 PI 0.00084839133 0.0028952067 Observations 184.00000 Degrees of Freedom 176.00000 Sum of Squared Errors 123.77150 Residual Variance 0.70325 R-squared 0.15257 Heteroskedasticity Test (P-Value) 0.14875 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.49216 95% Confidence Interval: [ 8.474243 , 8.504006 ] Sum of Squared Errors 72.01799 Residual Variance 0.42868 Joint R-Squared 0.50691 Heteroskedasticity Test (p-value) 0.00410 _____________________________________________ Regime 1: LRGDP0 <= 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.7912067 2.6456895 LM0 0.14281153 0.22274527 LS 0.27543552 0.15329479 LPRIMARY 0.057567688 0.28893203 LPOP 0.80857274 0.84765450 LRGDP0 0.48024014 0.17590938 NRP 1.6832783 0.31525026 PI 0.0052003173 0.0031656546 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -8.4122379 3.6267253 LM0 -0.30686785 0.57952714 LS -0.038131023 0.61021225 LPRIMARY -0.53792145 0.64183341 LPOP -1.2065894 2.4699756 LRGDP0 0.090048738 0.88438823 NRP 1.0050826 2.3130761 PI -0.0011676807 0.012201948 Observations 106.00000 Degrees of Freedom 98.00000 Sum of Squared Errors 66.83903 Residual Variance 0.68203 R-squared 0.39603 _____________________________________________ Regime 2: LRGDP0 > 8.492162 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 13.887127 2.5477955 LM0 -3.5515079 0.75811159 LS 0.54585406 0.15068760 LPRIMARY -0.81728657 1.4033883 LPOP -0.37535306 0.35330996 LRGDP0 -0.016775357 0.16362756 NRP -0.053236563 0.059240188 PI -0.0023170878 0.0020560693 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 5.4804001 21.310641 LM0 -5.4292549 0.081370448 LS 0.19685306 0.93867237 LPRIMARY -7.5373313 3.9497447 LPOP -1.0814176 1.6909179 LRGDP0 -0.61321594 0.32032486 NRP -0.18529207 0.12606820 PI -0.0068559897 0.0036427579 Observations 78.00000 Degrees of Freedom 70.00000 Sum of Squared Errors 5.17896 Residual Variance 0.07399 R-squared 0.73201 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 5.8198033 1.7748941 LM0 0.55274277 0.26381375 LS 0.26578408 0.17516247 LPRIMARY 0.41434997 0.32857407 LPOP 1.2471748 0.79609285 LRGDP0 -0.54771397 0.16746661 NRP 0.22160566 0.18797398 PI 0.00084839133 0.0028952067 Observations 184.00000 Degrees of Freedom 176.00000 Sum of Squared Errors 123.77150 Residual Variance 0.70325 R-squared 0.15257 Heteroskedasticity Test (P-Value) 0.14875 __________________________________________________ Threshold Estimation Threshold Variable = NRP Threshold Estimate 0.10150 95% Confidence Interval: [ 0.080100 , 0.102300 ] Sum of Squared Errors 89.79823 Residual Variance 0.53451 Joint R-Squared 0.38517 Heteroskedasticity Test (p-value) 0.08257 _____________________________________________ Regime 1: NRP <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 8.6934379 2.3496941 LM0 0.47839201 0.26611101 LS 0.095471520 0.15919945 LPRIMARY 0.098005913 0.32354059 LPOP 3.2227740 0.93048129 LRGDP0 -0.16762819 0.14411710 NRP 0.77368115 0.42156697 PI 0.00083596884 0.0029515546 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 3.7682176 13.510995 LM0 -0.053942628 1.0168015 LS -0.22947328 0.41267322 LPRIMARY -0.53613364 0.76138381 LPOP 1.3139005 5.0775739 LRGDP0 -0.47482336 0.12623102 NRP -0.052590119 1.7171773 PI -0.0052434151 0.0066847758 Observations 117.00000 Degrees of Freedom 109.00000 Sum of Squared Errors 72.24616 Residual Variance 0.66281 R-squared 0.14535 _____________________________________________ Regime 2: NRP > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 16.335956 3.3299183 LM0 -3.2890449 0.71564465 LS 1.1934123 0.32180535 LPRIMARY 3.7656437 0.59017596 LPOP -0.46780446 0.62881971 LRGDP0 -0.74885366 0.31055545 NRP 0.010192200 0.15426552 PI 0.0090360300 0.0055744574 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 6.6431629 24.644643 LM0 -4.9905130 -0.36073159 LS 0.47414897 1.8241508 LPRIMARY 2.2601685 5.2554569 LPOP -2.1552312 0.79446619 LRGDP0 -1.8508106 -0.12048985 NRP -0.29216822 0.35247284 PI -0.0024444942 0.023977881 Observations 67.00000 Degrees of Freedom 59.00000 Sum of Squared Errors 17.55207 Residual Variance 0.29749 R-squared 0.70884 __________________________________________________ F) Border market support: 1960 -- 1990 --------------------- Senstivity specification (1f) ---------------------------- ------- Baseline output growth regression (specification 5f in distortthr09.out) plus RRP as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34240 LM-test for no threshold 25.11588 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.40850 LM-test for no threshold 20.92964 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.015630456 0.042651216 LS 0.014169360 0.0036167017 LPRIMARY 0.016764512 0.0054915397 LPOP -0.044772064 0.019002557 LRGDP0 -0.014677257 0.0036512586 BMS 0.011727661 0.0049067308 Observations 152.00000 Degrees of Freedom 146.00000 Sum of Squared Errors 0.05745 Residual Variance 0.00039 R-squared 0.29768 Heteroskedasticity Test (P-Value) 0.00442 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.24237 95% Confidence Interval: [ 6.505933 , 8.890107 ] Sum of Squared Errors 0.04533 Residual Variance 0.00032 Joint R-Squared 0.44581 Heteroskedasticity Test (p-value) 0.02435 _____________________________________________ Regime 1: LRGDP0 <= 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.074109968 0.14795669 LS 0.012382647 0.0043538119 LPRIMARY 0.010504355 0.0083442176 LPOP -0.10117093 0.044982797 LRGDP0 -0.027434957 0.017898563 BMS 0.044920497 0.015684096 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.38056743 0.21588515 LS 0.0035753532 0.020916118 LPRIMARY -0.0058503119 0.026961767 LPOP -0.19200244 -0.013004648 LRGDP0 -0.062516140 0.0086003373 BMS 0.013400474 0.077506197 Observations 45.00000 Degrees of Freedom 39.00000 Sum of Squared Errors 0.01724 Residual Variance 0.00044 R-squared 0.27506 _____________________________________________ Regime 2: LRGDP0 > 7.242368 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.052211017 0.041050897 LS 0.017292148 0.0042350757 LPRIMARY 0.030490043 0.0080494900 LPOP -0.047295645 0.018633638 LRGDP0 -0.023882672 0.0036924640 BMS 0.010581452 0.0043415566 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.072963610 0.26493625 LS -0.0015758351 0.028299554 LPRIMARY -0.042203243 0.083183331 LPOP -0.083817575 0.027479190 LRGDP0 -0.031119901 -0.0091858353 BMS -7.0377587E-006 0.019880806 Observations 107.00000 Degrees of Freedom 101.00000 Sum of Squared Errors 0.02809 Residual Variance 0.00028 R-squared 0.42226 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.015630456 0.042651216 LS 0.014169360 0.0036167017 LPRIMARY 0.016764512 0.0054915397 LPOP -0.044772064 0.019002557 LRGDP0 -0.014677257 0.0036512586 BMS 0.011727661 0.0049067308 Observations 152.00000 Degrees of Freedom 146.00000 Sum of Squared Errors 0.05745 Residual Variance 0.00039 R-squared 0.29768 Heteroskedasticity Test (P-Value) 0.00442 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ -0.46510 , 0.596500 ] Sum of Squared Errors 0.04565 Residual Variance 0.00033 Joint R-Squared 0.44198 Heteroskedasticity Test (p-value) 0.00251 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.033214793 0.059999463 LS 0.012207746 0.0035448437 LPRIMARY 0.012532603 0.0053184213 LPOP -0.041624805 0.024019013 LRGDP0 -0.010675279 0.0038448070 BMS 0.031963661 0.014080114 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -1.0589475 0.084621075 LS 0.0052598527 0.028172647 LPRIMARY -0.033443604 0.024973384 LPOP -0.31984946 0.0054524614 LRGDP0 -0.020965402 0.038237965 BMS -0.0067452099 0.26664813 Observations 98.00000 Degrees of Freedom 92.00000 Sum of Squared Errors 0.03965 Residual Variance 0.00043 R-squared 0.22090 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.19839990 0.052937464 LS 0.026346605 0.0044826579 LPRIMARY 0.041607456 0.0074694667 LPOP -0.015242599 0.021117506 LRGDP0 -0.032151331 0.0039741751 BMS 0.0028383180 0.0030308172 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.083432698 0.45855184 LS 0.0022819433 0.041646814 LPRIMARY -0.011359626 0.15318556 LPOP -0.091002134 0.049659954 LRGDP0 -0.054545131 -0.012558673 BMS -0.0031020838 0.018127336 Observations 54.00000 Degrees of Freedom 48.00000 Sum of Squared Errors 0.00600 Residual Variance 0.00012 R-squared 0.70020 __________________________________________________ --------------------- Sensitivity specification (2f) --------------------------- -------- Baseline L_M growth regression (specification 6f in distortthr09.out) plus BMS as an additional independent variable Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 8.19420 LM-test for no threshold 32.66995 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.05330 LM-test for no threshold 25.35700 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.7681959 2.4756343 LM0 0.41761231 0.35296041 LS 0.16663866 0.15454024 LPRIMARY 0.63834484 0.42593348 LPOP 1.2790988 1.0644047 LRGDP0 -0.54112714 0.19507735 BMS 0.44875327 0.31432602 Observations 128.00000 Degrees of Freedom 121.00000 Sum of Squared Errors 113.72548 Residual Variance 0.93988 R-squared 0.12755 Heteroskedasticity Test (P-Value) 0.24583 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 63.19859 Residual Variance 0.55437 Joint R-Squared 0.51517 Heteroskedasticity Test (p-value) 0.10922 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.1892553 2.6118196 LM0 0.059553633 0.32246846 LS -0.012703249 0.13246372 LPRIMARY 0.22775327 0.38527451 LPOP 0.48461212 0.93910681 LRGDP0 0.49601869 0.20452728 BMS 1.9769920 0.32983221 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -6.3084217 5.2478155 LM0 -0.62421766 0.69159182 LS -0.27233214 0.26173991 LPRIMARY -0.52738478 1.0425382 LPOP -1.3560372 2.6540482 LRGDP0 0.018523771 0.89689217 BMS 1.3305209 2.7002068 Observations 76.00000 Degrees of Freedom 69.00000 Sum of Squared Errors 58.27830 Residual Variance 0.84461 R-squared 0.41333 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.490265 3.2813780 LM0 -4.2617179 0.93273514 LS 0.71866356 0.21904759 LPRIMARY -2.8904249 2.0360840 LPOP -0.045580644 0.55061279 LRGDP0 -0.045769999 0.21627335 BMS -0.090577347 0.086186426 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 8.3031143 24.530667 LM0 -6.0898788 -1.9616297 LS 0.20841488 1.1479968 LPRIMARY -6.8811495 3.0259558 LPOP -1.3703993 1.0336204 LRGDP0 -0.58522798 0.37812576 BMS -0.25950274 0.088412656 Observations 52.00000 Degrees of Freedom 45.00000 Sum of Squared Errors 4.92030 Residual Variance 0.10934 R-squared 0.76271 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.7681959 2.4756343 LM0 0.41761231 0.35296041 LS 0.16663866 0.15454024 LPRIMARY 0.63834484 0.42593348 LPOP 1.2790988 1.0644047 LRGDP0 -0.54112714 0.19507735 BMS 0.44875327 0.31432602 Observations 128.00000 Degrees of Freedom 121.00000 Sum of Squared Errors 113.72548 Residual Variance 0.93988 R-squared 0.12755 Heteroskedasticity Test (P-Value) 0.24583 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10230 95% Confidence Interval: [ 0.080100 , 0.108800 ] Sum of Squared Errors 78.31126 Residual Variance 0.68694 Joint R-Squared 0.39923 Heteroskedasticity Test (p-value) 0.17585 _____________________________________________ Regime 1: BMS <= 0.102300 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 10.385072 2.9405223 LM0 0.42403389 0.38963656 LS -0.045723992 0.13883698 LPRIMARY 0.18733682 0.43043455 LPOP 3.7912601 1.2096311 LRGDP0 -0.10126053 0.16940134 BMS 1.1982874 0.52630399 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 4.5522417 16.148496 LM0 -0.35538321 1.1877215 LS -0.31784447 0.22802366 LPRIMARY -0.69316890 1.0309885 LPOP 1.4203831 6.2363035 LRGDP0 -0.43328715 0.26612364 BMS 0.032425734 2.2298432 Observations 83.00000 Degrees of Freedom 76.00000 Sum of Squared Errors 65.33400 Residual Variance 0.85966 R-squared 0.15428 _____________________________________________ Regime 2: BMS > 0.102300 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 23.303227 4.5548233 LM0 -4.4361677 1.0684375 LS 0.73618628 0.32940604 LPRIMARY 4.7429473 0.74112875 LPOP 0.47611451 0.84243177 LRGDP0 -0.43778903 0.37367202 BMS 0.26422540 0.26514144 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 12.180199 32.230680 LM0 -6.5623034 -2.0690380 LS 0.090550434 1.5830736 LPRIMARY 2.8016791 6.1955597 LPOP -1.2425028 2.1272808 LRGDP0 -1.1701862 0.39586147 BMS -0.28774761 0.78390261 Observations 45.00000 Degrees of Freedom 38.00000 Sum of Squared Errors 12.97725 Residual Variance 0.34151 R-squared 0.75436 __________________________________________________ --------------------- Senstivity specification (5f) ---------------------------- ------- Baseline output growth regression (specification 5f in distortthr09.out) plus RRP and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.35021 LM-test for no threshold 24.08196 Bootstrap P-Value 0.00100 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.52830 LM-test for no threshold 18.93503 Bootstrap P-Value 0.01400 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.034252310 0.034641568 LS 0.011144042 0.0036564165 LPRIMARY 0.016513408 0.0052660178 LPOP -0.055776050 0.014681885 LRGDP0 -0.014860172 0.0032999359 BMS 0.0091978030 0.0046611915 PREMIUM -3.6664841E-005 1.2514140E-005 Observations 148.00000 Degrees of Freedom 141.00000 Sum of Squared Errors 0.04935 Residual Variance 0.00035 R-squared 0.33669 Heteroskedasticity Test (P-Value) 0.02767 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 7.66209 95% Confidence Interval: [ 7.505272 , 7.675453 ] Sum of Squared Errors 0.03909 Residual Variance 0.00029 Joint R-Squared 0.47461 Heteroskedasticity Test (p-value) 0.08027 _____________________________________________ Regime 1: LRGDP0 <= 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.15825559 0.10510345 LS 0.0092353568 0.0040283856 LPRIMARY 0.012992561 0.0061470326 LPOP -0.099835347 0.027913062 LRGDP0 -0.013013171 0.0098375254 BMS 0.034486182 0.014874266 PREMIUM -2.4406063E-005 9.3856759E-006 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.36549689 0.054920015 LS 0.0012168646 0.017456473 LPRIMARY -6.5599670E-005 0.025040745 LPOP -0.15902664 -0.038384360 LRGDP0 -0.036249259 0.0081343047 BMS 0.0024606796 0.064216572 PREMIUM -4.3592880E-005 -5.4489882E-006 Observations 64.00000 Degrees of Freedom 57.00000 Sum of Squared Errors 0.02261 Residual Variance 0.00040 R-squared 0.33590 _____________________________________________ Regime 2: LRGDP0 > 7.662091 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.063007195 0.044435293 LS 0.016707035 0.0043992087 LPRIMARY -0.016382915 0.026463571 LPOP -0.054311812 0.013899462 LRGDP0 -0.026527642 0.0041885183 BMS 0.0042490790 0.0035155388 PREMIUM -0.00035640596 0.00012560114 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.045546950 0.15010037 LS 0.0080845860 0.028289355 LPRIMARY -0.076275095 0.035485683 LPOP -0.085723046 -0.027001356 LRGDP0 -0.034775102 -0.017472441 BMS -0.0034705618 0.011139535 PREMIUM -0.00060258419 -8.4977945E-005 Observations 84.00000 Degrees of Freedom 77.00000 Sum of Squared Errors 0.01649 Residual Variance 0.00021 R-squared 0.48305 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.034252310 0.034641568 LS 0.011144042 0.0036564165 LPRIMARY 0.016513408 0.0052660178 LPOP -0.055776050 0.014681885 LRGDP0 -0.014860172 0.0032999359 BMS 0.0091978030 0.0046611915 PREMIUM -3.6664841E-005 1.2514140E-005 Observations 148.00000 Degrees of Freedom 141.00000 Sum of Squared Errors 0.04935 Residual Variance 0.00035 R-squared 0.33669 Heteroskedasticity Test (P-Value) 0.02767 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate -0.38760 95% Confidence Interval: [ -0.38760 , -0.37490 ] Sum of Squared Errors 0.04113 Residual Variance 0.00031 Joint R-Squared 0.44726 Heteroskedasticity Test (p-value) 0.46554 _____________________________________________ Regime 1: BMS <= -0.38760 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -1.4975434 0.52141344 LS 0.027702480 0.0066021642 LPRIMARY -0.015283651 0.0067655146 LPOP -0.49081430 0.18528873 LRGDP0 0.033661190 0.0081334426 BMS 0.10580722 0.053891922 PREMIUM -5.0936660E-005 2.2332371E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.5195137 -0.35859782 LS 0.012606974 0.040642722 LPRIMARY -0.028544059 -0.0020232420 LPOP -0.85398022 -0.084516637 LRGDP0 0.016578745 0.049602738 BMS -0.091631091 0.22562245 PREMIUM -9.4708107E-005 2.0577157E-006 Observations 10.00000 Degrees of Freedom 3.00000 Sum of Squared Errors 0.00082 Residual Variance 0.00027 R-squared 0.78160 _____________________________________________ Regime 2: BMS > -0.38760 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.012718710 0.035214670 LS 0.014702542 0.0036326725 LPRIMARY 0.020230289 0.0050473447 LPOP -0.055878945 0.014659397 LRGDP0 -0.018422779 0.0029118295 BMS 0.0074603910 0.0044036268 PREMIUM -4.3761641E-005 2.0584478E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.081739463 0.058721977 LS 0.0075825039 0.022805548 LPRIMARY 0.0098371136 0.030166842 LPOP -0.084611363 -0.025491985 LRGDP0 -0.024266360 -0.012715593 BMS -0.0011707175 0.016577574 PREMIUM -8.4107218E-005 -2.6303988E-006 Observations 138.00000 Degrees of Freedom 131.00000 Sum of Squared Errors 0.04030 Residual Variance 0.00031 R-squared 0.36580 __________________________________________________ --------------------- Sensitivity specification (6f) --------------------------- -------- Baseline L_M growth regression (specification 6f in distortthr09.out) plus BMS and premium as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 8.19420 LM-test for no threshold 32.10154 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.01300 LM-test for no threshold 24.44856 Bootstrap P-Value 0.00100 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 8.1963752 2.7159992 LM0 0.37557331 0.34879381 LS 0.068210795 0.16720260 LPRIMARY 0.68864536 0.42336219 LPOP 1.7994068 1.1777839 LRGDP0 -0.48699096 0.19816518 BMS 0.55730530 0.34434290 PREMIUM -0.00099967562 0.00059321784 Observations 125.00000 Degrees of Freedom 117.00000 Sum of Squared Errors 108.31459 Residual Variance 0.92577 R-squared 0.15028 Heteroskedasticity Test (P-Value) 0.32358 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.492162 ] Sum of Squared Errors 62.75748 Residual Variance 0.57576 Joint R-Squared 0.50768 Heteroskedasticity Test (p-value) 0.13107 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.65395435 3.2233369 LM0 0.059974726 0.32947072 LS -0.028775242 0.15029444 LPRIMARY 0.23895848 0.39388718 LPOP 0.66229547 1.1884778 LRGDP0 0.49228907 0.20998944 BMS 1.9673070 0.33896638 PREMIUM -0.00015514675 0.00034202677 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -6.9716947 7.2361878 LM0 -0.63531526 0.70573734 LS -0.32335235 0.27757798 LPRIMARY -0.53306039 1.0724533 LPOP -1.6671209 3.4293342 LRGDP0 0.0047497819 0.90386838 BMS 1.3029328 2.7052085 PREMIUM -0.00088673886 0.00052497303 Observations 74.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 58.13097 Residual Variance 0.88077 R-squared 0.40788 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 18.618834 3.2730330 LM0 -4.3260015 0.96642108 LS 0.67246834 0.21916756 LPRIMARY -2.5501412 1.8440410 LPOP 0.17514026 0.55225144 LRGDP0 -0.052368373 0.23905872 BMS -0.048546430 0.080897746 PREMIUM -0.0024925597 0.0025860688 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 9.2458337 25.240268 LM0 -6.2201868 -2.0628492 LS 0.19755852 1.1020368 LPRIMARY -6.1644616 2.9684766 LPOP -1.1366947 1.2575531 LRGDP0 -0.56417414 0.41618671 BMS -0.20710601 0.12480194 PREMIUM -0.0077564012 0.0067873985 Observations 51.00000 Degrees of Freedom 43.00000 Sum of Squared Errors 4.62651 Residual Variance 0.10759 R-squared 0.77220 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 8.1963752 2.7159992 LM0 0.37557331 0.34879381 LS 0.068210795 0.16720260 LPRIMARY 0.68864536 0.42336219 LPOP 1.7994068 1.1777839 LRGDP0 -0.48699096 0.19816518 BMS 0.55730530 0.34434290 PREMIUM -0.00099967562 0.00059321784 Observations 125.00000 Degrees of Freedom 117.00000 Sum of Squared Errors 108.31459 Residual Variance 0.92577 R-squared 0.15028 Heteroskedasticity Test (P-Value) 0.32358 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.101500 , 0.346900 ] Sum of Squared Errors 75.93763 Residual Variance 0.69668 Joint R-Squared 0.40428 Heteroskedasticity Test (p-value) 0.23909 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 10.339745 3.0634814 LM0 0.44416126 0.39460657 LS -0.10031916 0.15168242 LPRIMARY 0.17066319 0.43190770 LPOP 3.7738359 1.2440570 LRGDP0 -0.093977498 0.17370383 BMS 0.93114757 0.55760835 PREMIUM -0.00058965764 0.00055173997 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 4.3353213 16.411106 LM0 -0.32926762 1.2357776 LS -0.40377735 0.19697839 LPRIMARY -0.67587590 1.0566369 LPOP 1.2446117 6.2121876 LRGDP0 -0.47484197 0.24648201 BMS -0.16176479 2.1418284 PREMIUM -0.0017027351 0.00049409496 Observations 81.00000 Degrees of Freedom 73.00000 Sum of Squared Errors 63.61476 Residual Variance 0.87144 R-squared 0.14769 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 23.664950 4.4626048 LM0 -3.8796288 1.2254768 LS 1.1755427 0.41646852 LPRIMARY 3.1543035 0.87293302 LPOP 1.1721394 0.96973513 LRGDP0 -0.67724045 0.52337963 BMS 0.28751061 0.26047478 PREMIUM -0.030417086 0.016076217 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 14.918245 33.807976 LM0 -6.3427547 -1.4776943 LS 0.19497537 1.9918210 LPRIMARY 1.4433548 5.8630868 LPOP -0.76686320 3.1119976 LRGDP0 -1.7030645 0.34858362 BMS -0.22301996 0.82441679 PREMIUM -0.061926472 0.0098573085 Observations 44.00000 Degrees of Freedom 36.00000 Sum of Squared Errors 12.32287 Residual Variance 0.34230 R-squared 0.76653 __________________________________________________ --------------------- Senstivity specification (7f) ---------------------------- ------- Baseline output growth regression (specification 5f in distortthr09.out) plus BMS and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.34780 LM-test for no threshold 25.36908 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.52830 LM-test for no threshold 20.60844 Bootstrap P-Value 0.00800 Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.0084852738 0.042133067 LS 0.012279457 0.0039572455 LPRIMARY 0.016101888 0.0054040964 LPOP -0.044453496 0.019137435 LRGDP0 -0.014533878 0.0036932760 BMS 0.012470436 0.0049247037 PI -2.6106818E-005 1.7857461E-005 Observations 150.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.05640 Residual Variance 0.00039 R-squared 0.30410 Heteroskedasticity Test (P-Value) 0.00933 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 6.67935 95% Confidence Interval: [ 6.639614 , 8.905376 ] Sum of Squared Errors 0.04468 Residual Variance 0.00033 Joint R-Squared 0.44871 Heteroskedasticity Test (p-value) 0.34098 _____________________________________________ Regime 1: LRGDP0 <= 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.71533906 0.24398830 LS 0.0055637625 0.0069344288 LPRIMARY -0.0037600105 0.0083708406 LPOP -0.072051383 0.063808650 LRGDP0 -0.13700071 0.028720086 BMS 0.092131076 0.018265977 PI -3.8510256E-005 2.0378645E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.36354574 1.4717527 LS -0.0094614111 0.025734817 LPRIMARY -0.030915804 0.047153461 LPOP -0.32880430 0.26833648 LRGDP0 -0.22664941 0.0079636265 BMS 0.011201930 0.15467239 PI -9.1019939E-005 3.8469211E-005 Observations 11.00000 Degrees of Freedom 4.00000 Sum of Squared Errors 0.00102 Residual Variance 0.00025 R-squared 0.80700 _____________________________________________ Regime 2: LRGDP0 > 6.679347 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.0078739270 0.042758825 LS 0.014866784 0.0035608757 LPRIMARY 0.019393696 0.0064305260 LPOP -0.043885527 0.018500295 LRGDP0 -0.016024862 0.0031603239 BMS 0.011600306 0.0047413269 PI -0.00013820244 5.5750402E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.081865870 0.29334938 LS 0.0020662253 0.030016594 LPRIMARY -0.047150777 0.10174340 LPOP -0.087768644 0.021661249 LRGDP0 -0.040504153 -0.0082576371 BMS -0.0014627224 0.022263386 PI -0.00031935329 0.00052284497 Observations 139.00000 Degrees of Freedom 132.00000 Sum of Squared Errors 0.04366 Residual Variance 0.00033 R-squared 0.40495 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: GGDP Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant -0.0084852738 0.042133067 LS 0.012279457 0.0039572455 LPRIMARY 0.016101888 0.0054040964 LPOP -0.044453496 0.019137435 LRGDP0 -0.014533878 0.0036932760 BMS 0.012470436 0.0049247037 PI -2.6106818E-005 1.7857461E-005 Observations 150.00000 Degrees of Freedom 143.00000 Sum of Squared Errors 0.05640 Residual Variance 0.00039 R-squared 0.30410 Heteroskedasticity Test (P-Value) 0.00933 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ -0.42340 , 0.596500 ] Sum of Squared Errors 0.04517 Residual Variance 0.00033 Joint R-Squared 0.44268 Heteroskedasticity Test (p-value) 0.00426 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -0.022637258 0.060290361 LS 0.010738784 0.0037475729 LPRIMARY 0.012154799 0.0053464568 LPOP -0.039450101 0.024509588 LRGDP0 -0.010594440 0.0038889852 BMS 0.031937726 0.014319702 PI -1.8998401E-005 1.8107208E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -2.6090645 0.56267839 LS -0.017704478 0.071850964 LPRIMARY -0.048589638 0.034405823 LPOP -0.60433904 0.11177489 LRGDP0 -0.027744099 0.10132573 BMS -0.23540188 0.39323279 PI -0.00083795171 0.0011811464 Observations 96.00000 Degrees of Freedom 89.00000 Sum of Squared Errors 0.03917 Residual Variance 0.00044 R-squared 0.22498 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 0.19840164 0.052815393 LS 0.026785878 0.0061748831 LPRIMARY 0.041926932 0.0077919346 LPOP -0.015386166 0.021079192 LRGDP0 -0.032425733 0.0049214655 BMS 0.0026836613 0.0030735564 PI 1.0115711E-005 8.2715698E-005 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -0.10712324 0.47108453 LS 0.0025719403 0.045059864 LPRIMARY -0.011736160 0.13077238 LPOP -0.10096336 0.056557426 LRGDP0 -0.060334335 -0.010879346 BMS -0.0033405093 0.018300604 PI -0.00026166062 0.00030700476 Observations 54.00000 Degrees of Freedom 47.00000 Sum of Squared Errors 0.00600 Residual Variance 0.00013 R-squared 0.70028 __________________________________________________ --------------------- Senstivity specification (8f) ---------------------------- ------- Baseline L_M growth regression (specification 6f in distortthr09.out) plus BMS and relative price of investment as additional independent variables Testing for a First Sample Split, Using Output Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 9.10981 LM-test for no threshold 29.87886 Bootstrap P-Value 0.00000 Testing for a First Sample Split, Using BMS Test of Null of No Threshold Against Alternative of Threshold Allowing Heteroskedastic Errors (White Corrected) Number of Bootstrap Replications 1000.00000 Trimming Percentage 0.15000 Threshold Estimate 0.03620 LM-test for no threshold 29.84330 Bootstrap P-Value 0.00000 Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.5863926 2.5216944 LM0 0.43365042 0.35827486 LS 0.20787850 0.21854814 LPRIMARY 0.61972481 0.45128095 LPOP 1.2349209 1.0459655 LRGDP0 -0.56087301 0.20367879 BMS 0.42652288 0.32411353 PI 0.00062388698 0.0034351884 Observations 126.00000 Degrees of Freedom 118.00000 Sum of Squared Errors 113.14543 Residual Variance 0.95886 R-squared 0.12962 Heteroskedasticity Test (P-Value) 0.20176 __________________________________________________ Threshold Estimation Threshold Variable = LRGDP0 Threshold Estimate 8.47424 95% Confidence Interval: [ 8.474243 , 8.474243 ] Sum of Squared Errors 57.18629 Residual Variance 0.51988 Joint R-Squared 0.56009 Heteroskedasticity Test (p-value) 0.02902 _____________________________________________ Regime 1: LRGDP0 <= 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant -5.1689420 3.2696882 LM0 9.5937440E-005 0.28524676 LS 0.33684502 0.21101182 LPRIMARY 0.088968882 0.40317684 LPOP -0.026658882 0.85523762 LRGDP0 0.68484732 0.21210310 BMS 1.7747736 0.32076652 PI 0.0063408651 0.0036542003 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant -11.577531 1.2396469 LM0 -0.55898772 0.55917959 LS -0.076738154 0.75042819 LPRIMARY -0.70125772 0.87919549 LPOP -1.7029246 1.6496069 LRGDP0 0.26912525 1.1005694 BMS 1.1460713 2.4034760 PI -0.00082136756 0.013503098 Observations 74.00000 Degrees of Freedom 66.00000 Sum of Squared Errors 52.45507 Residual Variance 0.79477 R-squared 0.46803 _____________________________________________ Regime 2: LRGDP0 > 8.474243 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 17.258121 3.2650760 LM0 -4.2415292 0.92479364 LS 0.58194948 0.20702933 LPRIMARY -3.1363976 1.9962709 LPOP 0.066384375 0.55773897 LRGDP0 0.089709633 0.22825164 BMS -0.040953703 0.088929706 PI -0.0048173223 0.0025331205 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 10.858572 23.657669 LM0 -6.0541247 -2.4289336 LS 0.17617199 0.98772696 LPRIMARY -7.0490884 0.77629330 LPOP -1.0267840 1.1595528 LRGDP0 -0.35766357 0.53708284 BMS -0.21525593 0.13334852 PI -0.0097822384 0.00014759380 Observations 52.00000 Degrees of Freedom 44.00000 Sum of Squared Errors 4.73122 Residual Variance 0.10753 R-squared 0.77183 __________________________________________________ Global OLS Estimation, Without Threshold Dependent Variable: LMHAT Heteroskedasticity Correction Used Variable Estimate St Error __________________________________________ Constant 6.5863926 2.5216944 LM0 0.43365042 0.35827486 LS 0.20787850 0.21854814 LPRIMARY 0.61972481 0.45128095 LPOP 1.2349209 1.0459655 LRGDP0 -0.56087301 0.20367879 BMS 0.42652288 0.32411353 PI 0.00062388698 0.0034351884 Observations 126.00000 Degrees of Freedom 118.00000 Sum of Squared Errors 113.14543 Residual Variance 0.95886 R-squared 0.12962 Heteroskedasticity Test (P-Value) 0.20176 __________________________________________________ Threshold Estimation Threshold Variable = BMS Threshold Estimate 0.10150 95% Confidence Interval: [ 0.080100 , 0.108800 ] Sum of Squared Errors 77.51881 Residual Variance 0.70472 Joint R-Squared 0.40368 Heteroskedasticity Test (p-value) 0.17501 _____________________________________________ Regime 1: BMS <= 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 9.8624468 3.0971220 LM0 0.40342462 0.38259015 LS 0.015290243 0.19783826 LPRIMARY 0.14488894 0.45040858 LPOP 3.7386117 1.1856841 LRGDP0 -0.077735362 0.17308004 BMS 1.0449209 0.52539353 PI 0.00094445912 0.0033488733 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 3.7920878 16.110033 LM0 -0.34645208 1.1838978 LS -0.39715588 0.40305322 LPRIMARY -0.73791188 1.0705668 LPOP 1.3672539 6.0625525 LRGDP0 -0.46081121 0.26150151 BMS 0.015149637 2.2250388 PI -0.0059868478 0.0075082509 Observations 80.00000 Degrees of Freedom 72.00000 Sum of Squared Errors 63.87778 Residual Variance 0.88719 R-squared 0.14425 _____________________________________________ Regime 2: BMS > 0.101500 Parameter Estimates Variable Estimate St Error ________________________________________ Constant 20.032404 3.6482810 LM0 -3.5544403 0.90369657 LS 1.2230968 0.36607978 LPRIMARY 4.2540611 0.57528509 LPOP 0.32445309 0.69376987 LRGDP0 -0.78152534 0.35175154 BMS 0.12085241 0.29914063 PI 0.0078614107 0.0059356375 95% Confidence Regions for Parameters Variable Low High ________________________________________ Constant 12.881773 31.288888 LM0 -5.9718890 -1.7831950 LS 0.38521283 1.9406132 LPRIMARY 3.1265024 6.0390457 LPOP -1.0761519 2.0733492 LRGDP0 -1.4709584 -0.0036736867 BMS -0.46546323 0.81837285 PI -0.0072979177 0.019495260 Observations 46.00000 Degrees of Freedom 38.00000 Sum of Squared Errors 13.64103 Residual Variance 0.35897 R-squared 0.75321 __________________________________________________