Table 3: Weighted Least Squares meta regression results
|
(1) T-stat |
(2) T-stat |
(3) T-stat |
(4) T-stat |
(5) T-stat |
(6) T-stat |
(7) T-stat | |
|
Constant |
2.693*** |
1.932*** |
-0.253* |
-0.572*** |
-0.406*** |
-28.82*** |
-0.360*** |
|
(0.077) |
(0.079) |
(0.101) |
(0.097) |
(0.109) |
(0.001) |
(0.006) | |
|
no. of explanatory variables |
0.0422*** |
0.0416*** |
0.0416*** |
0.0416*** |
0.0425*** | ||
|
(0.000) |
(0.000) |
(0.000) |
(0.000) |
(0.000) | |||
|
N (in logs) |
0.0175*** |
0.000443 |
0.00515 |
-0.00195 |
-0.0162*** | ||
|
(0.000) |
(0.006) |
(0.005) |
(0.006) |
(0.000) | |||
|
No. of years used (log) |
0.0448*** |
-0.392* |
0.0198 |
-0.269 |
-0.378*** | ||
|
(0.002) |
(0.166) |
(0.086) |
(0.375) |
(0.025) | |||
|
No. of countries (log) |
0.384*** |
-0.0535*** |
-0.0642*** |
-0.0561*** |
-0.0285*** | ||
|
(0.003) |
(0.010) |
(0.010) |
(0.010) |
(0.000) | |||
|
annual = 1 |
1.241*** |
-2.880* |
1.092*** | ||||
|
(0.190) |
(1.214) |
(0.031) | |||||
|
selection correction = 1 |
-0.190*** |
-0.189*** |
-0.252*** | ||||
|
(0.049) |
(0.049) |
(0.004) | |||||
|
published = 1 |
0.0106 |
2.287* |
0.0117*** | ||||
|
(0.009) |
(0.977) |
(0.000) | |||||
|
country effects = 1 |
-0.283* |
0.384 |
-0.235*** | ||||
|
(0.115) |
(0.560) |
(0.016) | |||||
|
time effects = 1 |
0.00382 |
0.0105 |
-0.00122 | ||||
|
(0.046) |
(0.048) |
(0.004) | |||||
|
cross section = 1 |
-2.256*** |
-4.005 |
-2.485*** | ||||
|
(0.320) |
(4.257) |
(0.044) | |||||
|
other preferences included = 1 |
0.200*** |
0.699 |
0.151*** | ||||
|
(0.061) |
(3.553) |
(0.010) | |||||
|
gravity regression = 1 |
-0.466*** |
4.300** |
-0.379*** | ||||
|
(0.066) |
(1.514) |
(0.011) | |||||
|
agoa dummy = 1 |
-2.707*** |
-2.196*** |
-2.673*** | ||||
|
(0.262) |
(0.422) |
(0.103) | |||||
|
robust s.e. = 1 |
-0.144*** |
-0.139*** |
0.00619*** | ||||
|
(0.016) |
(0.017) |
(0.001) | |||||
|
product effects = 1 |
0.153** |
-0.864 |
0.391*** | ||||
|
(0.047) |
(0.570) |
(0.005) | |||||
|
single country analysis = 1 |
0.834*** |
0.972*** | |||||
|
(0.125) |
(0.015) | ||||||
|
logged dep. var. = 1 |
0.288*** |
-0.116 |
0.331*** | ||||
|
(0.041) |
(0.208) |
(0.005) | |||||
|
product group = ”All/Total” |
-0.00637 |
-0.0311 | |||||
|
(0.056) |
(0.068) | ||||||
|
product group = ”Apparel & Textiles” |
-0.0369 |
-0.0436 | |||||
|
(0.045) |
(0.045) | ||||||
|
product group = ”Agriculture” |
0.0609* |
0.0550* | |||||
|
(0.026) |
(0.026) | ||||||
|
regions included = AGOA countries |
-0.0514 |
-2.874 | |||||
|
(0.068) |
(3.493) | ||||||
|
regions included = AGOA + N. Africa |
-0.287*** |
-6.300** | |||||
|
(0.069) |
(2.054) | ||||||
|
definition of dep. var. = Exports |
-0.214 |
0.696 | |||||
|
(0.264) |
(3.502) | ||||||
|
definition of dep. var. = Imports |
0.156 |
1.813 | |||||
|
(0.245) |
(3.455) | ||||||
|
time frequency = Annual |
0.488** | ||||||
|
(0.157) | |||||||
|
time frequency = Monthly |
-0.0248 |
-2.959* | |||||
|
(0.134) |
(1.208) | ||||||
|
econometric method = GMM |
-0.00367 |
-0.657 | |||||
|
(0.057) |
(0.607) | ||||||
|
econometric method = Heckman/Tobit |
-0.873*** |
-2.552 | |||||
|
(0.121) |
(1.567) | ||||||
|
level of aggregation = Aggregate |
0.0686+ |
0.0712+ | |||||
|
(0.039) |
(0.040) | ||||||
|
level of aggregation = Less Aggregate |
0.0779* |
-2.206* | |||||
|
(0.040) |
(0.976) | ||||||
|
Precision |
0.00449*** |
-2.228*** |
2.584*** |
-0.258+ |
4.384*** |
0.0128*** |
2.489*** |
|
(0.000) |
(0.009) |
(0.342) |
(0.149) |
(1.056) |
(0.000) |
(0.107) | |
|
Observations |
173 |
173 |
173 |
173 |
173 |
173 |
173 |
|
goodness of fit-χ2 |
120496.0 |
7710.6 |
862.6 |
1074.5 |
836.8 |
1.31107e+10 |
316374.3 |
|
model χ2 |
1217.4 |
114002.8 |
120850.8 |
120639.0 |
120876.6 |
600194512.9 |
1.37106e+10 |
|
Standard errors in parentheses. All explanatory variables divided by standard error. |
The last two columns use the square of the standard error as weights | ||||||
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001
Multi-Level Meta-Regression
The final table, Table (4) below presents the multi-level MRA results. We include these results to check
the robustness of our earlier results given that that our coefficients are nested in the individual studies. The
intercept is the only random component included in the 2-level multi-level MRA. In column (2) in the table,
precision is negative and significant at the 1% level of significance. The number of explanatory variables,
product fixed effects and logged dependent variables contribute to higher to t-statistics holding all else con-
stant. The remaining significant variables in the final column of the table, number of years, cross-sectional
14
More intriguing information
1. Flatliners: Ideology and Rational Learning in the Diffusion of the Flat Tax2. Place of Work and Place of Residence: Informal Hiring Networks and Labor Market Outcomes
3. A dynamic approach to the tendency of industries to cluster
4. Governance Control Mechanisms in Portuguese Agricultural Credit Cooperatives
5. The name is absent
6. The name is absent
7. Quality Enhancement for E-Learning Courses: The Role of Student Feedback
8. Estimation of marginal abatement costs for undesirable outputs in India's power generation sector: An output distance function approach.
9. THE AUTONOMOUS SYSTEMS LABORATORY
10. The name is absent