Table 2.
Regression results
| Variables | GHES/GDP |
|---|---|
| Lag GHES/GDP | 0.924*** (0.062) |
| DAHG/GDP | −0.620** (0.287) |
| DAHG-/GDP | 0.865* (0.497) |
| DAHNG/GDP | 0.442* (0.245) |
| GGE/GDP | 0.008* (0.004) |
| Ln GDP per cap | 0.001 (0.001) |
| GDP growth rate | 0.017*** (0.004) |
| Ln population | 0.000 (0.000) |
| Constant | −0.001 (0.007) |
| Observations | 1762 |
| Number of countries | 119 |
| Effect of decrease in DAHG (beta estimate) | 0.245 |
| Effect of decrease in DAHG (standard error) | (0.562) |
| Test up = down p-value | 0.082 |
| Hansen p-value | 0.272 |
| AR(2) p-value | 0.345 |
| Instrument count | 37 |
Two-step system GMM estimation of model (2). Coefficient estimates for time indicators suppressed. DAHG, DAHNG, and LDV treated as endogenous. Instrument matrix is ‘collapsed’ and factored using principal components analysis. Components with eigenvalue greater than one retained. Windmeijer standard errors reported. Forward orthogonal deviation used for the transformed equation, rather than first-differencing. Backward orthogonal deviation used for the instruments of the level equation, rather than first-differencing.
Windmeijer adjusted standard errors in parentheses.
***p < 0.01,
**p < 0.05,
*p < 0.1.