Table 11: Budget balances and structural reforms: results from the estimation of fiscal rules (sample defined
in such a way that there are no missing observations for any of the reform dummies)
|
Dependent variable: |
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
(7) |
(8) |
|
Explanatory variables | ||||||||
|
Constant |
-2.2*** |
-2.2*** |
-2.58*** |
-2.67*** |
-2.35*** |
-2.39*** |
-2.62*** |
-2.57*** |
|
(-3.65) |
(-3.65) |
(-4.14) |
(-4.26 |
(-3.78) |
(-4.01) |
(-4.12) |
(-4.14) | |
|
Lagged dependent |
0.75*** |
0.76*** |
0.75*** |
0.74*** |
0.76*** |
0.75*** |
0.77*** |
0.75*** |
|
variable |
(16.26) |
(16.58) |
(16.66) |
(16.02) |
(16.6) |
(16.79) |
(16.99) |
(16.68) |
|
Output gap |
0.019 |
0.006 |
0.018 |
0.034 |
0.012 |
0.022 |
0.005 |
0.016 |
|
(0.42) |
(0.14) |
(0.41) |
(0.73) |
(0.28) |
(0.5) |
(0.13) |
(0.36) | |
|
Lagged debt/GDP ratio |
0.044*** |
0.043*** |
0.051*** |
0.055*** |
0.04*** |
0.047*** |
0.05*** |
0.049*** |
|
(4.73) |
(4.6) |
(5.18) |
(5.33) |
(4.71) |
(5.19) |
(5.04) |
(5.16) | |
|
Dummy labour market |
-0.26 |
-0.34 | ||||||
|
reforms |
(-1.18) |
(-1.55) | ||||||
|
Dummy product market |
-0.12 |
0.01 | ||||||
|
reforms |
(-0.56) |
(0.05) | ||||||
|
Dummy pension |
-0.49** |
-0.54** | ||||||
|
reforms Interaction labour |
(-2.26) |
(-2.39) |
-0.24 | |||||
|
market/product |
(-1.1) | |||||||
|
Interaction labour |
-0.68*** | |||||||
|
market/pension reforms Interaction product |
(-2.74) |
-0.37* | ||||||
|
market/pension |
(-1.68) | |||||||
|
Interaction labour |
-0.65** | |||||||
|
market/product market/pension reforms |
(-2.38) | |||||||
|
N. obs. |
188 |
188 |
188 |
188 |
188 |
188 |
188 |
188 |
|
R sq. |
0.65 |
0.65 |
0.65 |
0.66 |
0.65 |
0.66 |
0.66 |
0.66 |
|
Chi sq |
853 |
848 |
875 |
878 |
853 |
941 |
908 |
879 |
Notes: Estimations method: fixed effects, instrumental variables regression. The output gap is instrumented with its own lag and
the US lagged output gap. All fiscal variables are expressed as shares on potential output.
Z statistics are reported in parentheses. *, **, and *** denote, respectively, significance at the 10, 5 and 1
per cent level. Coefficients for country fixed effects are not reported.
The pension reform dummy is constructed as an indicator taking value 1 if a pension reform was carried out in the current or
previous year and zero otherwise.
Source: Authors’ calculation on data described in table 1 and DG ECFIN AMECO database.
129