Table 1: Impact of tax reform on economic growth
Estimator: Two-step System GMM
Variables
|
GROWTH
|
GROWTH
|
GROWTH
|
GROWTH
|
|
(1)
|
(2)
|
(3)
|
(4)
|
GROWTHt-1
|
0.0457***
|
0.0540***
|
0.0485***
|
0.0468***
|
|
(0.00860)
|
(0.00725)
|
(0.00710)
|
(0.00802)
|
TAXREF
|
0.0758***
|
0.0994***
|
0.0536***
|
0.0639***
|
|
(0.00812)
|
(0.0103)
|
(0.0105)
|
(0.00638)
|
LIC*TAXREF
|
|
-0.0866***
|
|
|
|
|
(0.0134)
|
|
|
LMIC*TAXREF
|
|
|
0.0121
|
|
|
|
|
(0.0107)
|
|
UMIC*TAXREF
|
|
|
|
0.0260*
|
|
|
|
|
(0.0134)
|
LIC
|
|
3.178***
|
|
|
|
|
(0.850)
|
|
|
LMIC
|
|
|
-1.251*
|
|
|
|
|
(0.723)
|
|
UMIC
|
|
|
|
-0.349
|
|
|
|
|
(0.797)
|
GFCF
|
0.139***
|
0.113***
|
0.144***
|
0.127***
|
|
(0.0121)
|
(0.00932)
|
(0.00935)
|
(0.0111)
|
EDU
|
0.00737
|
-0.0166**
|
0.0134***
|
0.000519
|
|
(0.00776)
|
(0.00770)
|
(0.00457)
|
(0.00531)
|
FINDEV
|
-0.0426***
|
-0.0525***
|
-0.0398***
|
-0.0413***
|
|
(0.00431)
|
(0.00443)
|
(0.00305)
|
(0.00234)
|
INFL
|
-0.597***
|
-0.547***
|
-0.581***
|
-0.564***
|
|
(0.0949)
|
(0.0738)
|
(0.0677)
|
(0.0605)
|
POLITY2
|
-0.0310
|
-0.0230
|
-0.0634***
|
-0.0140
|
|
(0.0275)
|
(0.0216)
|
(0.0172)
|
(0.0193)
|
Log(POP)
|
0.908***
|
1.096***
|
0.759***
|
0.600***
|
|
(0.102)
|
(0.118)
|
(0.0719)
|
(0.0777)
|
Constant
|
-18.65***
|
-20.96***
|
-15.32***
|
-13.00***
|
|
(1.511)
|
(1.750)
|
(1.070)
|
(1.137)
|
|
|
|
|
|
Observations - Countries
|
398 - 92
|
398 - 92
|
398 - 92
|
398 - 92
|
Number of Instruments
|
84
|
91
|
91
|
91
|
AR1 (P-Value)
|
0.0006
|
0.0007
|
0.0005
|
0.0006
|
AR2 (P-Value)
|
0.9366
|
0.9122
|
0.9278
|
0.8741
|
AR3 (P-Value)
|
0.4279
|
0.7186
|
0.4275
|
0.5424
|
OID (P-Value)
|
0.2674
|
0.5070
|
0.3623
|
0.2054
|
Note: *p-value<0.1; **p-value<0.05; ***p-value<0.01. Robust Standard Errors are in parenthesis. In the two-step system GMM estimations with the Windmeijer (2005) correction for finite sample bias, the variables "TAXREF", "EDU", "INFL", "GFCF", "FINDEV", "POLITY2" and the interaction variables have been considered as endogenous. The variable "POP" has been considered as exogenous. Time dummies have been included in the regressions.
Table 2: Impact of tax reform on economic growth through public revenue and trade openness
Estimator: Two-step System GMM
Variables
|
GROWTH
|
GROWTH
|
GROWTH
|
|
(1)
|
(2)
|
(3)
|
GROWTHt-1
|
0.0392***
|
0.0491***
|
0.0413***
|
|
(0.00493)
|
(0.00616)
|
(0.00524)
|
TAXREF
|
0.147***
|
-0.0382***
|
-0.0702***
|
|
(0.0187)
|
(0.0128)
|
(0.0118)
|
TAXREF*[Log(OPENSW)]
|
0.0138***
|
|
|
|
(0.00188)
|
|
|
Log(OPENSW)
|
-0.552***
|
|
|
|
(0.137)
|
|
|
TAXREF*OPEN
|
|
0.00103***
|
|
|
|
(0.000207)
|
|
OPEN
|
|
-0.0350***
|
|
|
|
(0.0129)
|
|
TAXREF*REVENUE
|
|
|
0.00560***
|
|
|
|
(0.000500)
|
REVENUE
|
|
|
-0.196***
|
|
|
|
(0.0184)
|
GFCF
|
0.0967***
|
0.0956***
|
0.158***
|
|
(0.0102)
|
(0.0103)
|
(0.00919)
|
EDU
|
0.00386
|
0.0190***
|
-0.0259***
|
|
(0.00640)
|
(0.00555)
|
(0.00629)
|
FINDEV
|
-0.0367***
|
-0.0578***
|
-0.0446***
|
|
(0.00477)
|
(0.00347)
|
(0.00283)
|
INFL
|
-1.186***
|
-0.982***
|
-0.998***
|
|
(0.0618)
|
(0.0672)
|
(0.0633)
|
POLITY2
|
0.00120
|
-0.0552***
|
0.0581***
|
|
(0.0156)
|
(0.0191)
|
(0.0127)
|
Log(POP)
|
0.399***
|
1.334***
|
0.674***
|
|
(0.122)
|
(0.0823)
|
(0.0897)
|
Constant
|
-10.36***
|
-19.00***
|
-6.704***
|
|
(3.116)
|
(1.367)
|
(1.704)
|
|
|
|
|
Observations - Countries
|
397 - 92
|
397 - 92
|
382 - 91
|
Number of Instruments
|
92
|
92
|
92
|
AR1 (P-Value)
|
0.0005
|
0.0004
|
0.0006
|
AR2 (P-Value)
|
0.8990
|
0.8252
|
0.7677
|
AR3 (P-Value)
|
0.3145
|
0.2784
|
0.4485
|
OID (P-Value)
|
0.2849
|
0.4010
|
0.3577
|
Note: *p-value<0.1; **p-value<0.05; ***p-value<0.01. Robust Standard Errors are in parenthesis. In the two-step system GMM estimations with the Windmeijer (2005) correction for finite sample bias, the variables "TAXREF", "OPENSW", "OPEN", "EDU", "INFL", "GFCF", "FINDEV", "POLITY2" and the interaction variables have been considered as endogenous. The variable "POP" has been considered as exogenous. Time dummies have been included in the regressions.