Table 4
Political outcomes in panel - periods correspond to legislatures
GMM estimates
(1) |
(2) |
(3) |
(4) |
(5) | |
Dep. Var. |
party_frag |
coalition |
single |
coalition |
single |
party_frag |
1.35 |
-1.26 | |||
(0.81)* |
(0.63)** | ||||
maj |
-0.06 |
0.37 |
0.14 | ||
(0.04) |
(0.48) |
(0.09)* | |||
district |
2.31 |
6.31 |
-0.10 | ||
(0.24)*** |
(3.02)** |
(0.56) | |||
threshold |
-0.00 |
0.16 |
-0.00 | ||
(0.00) |
(0.10) |
(0.01) | |||
Over-id |
4.16 (3) |
0.01 (1) |
2.48 (1) |
3.83 (6) |
3.70 (6) |
AR(2) |
0.68 |
-0.02 |
0.93 |
0.01 |
0.46 |
Estimation |
GMM |
GMM |
GMM |
GMM |
GMM |
Obs. |
141 |
97 |
97 |
97 |
97 |
N. countries |
37 |
35 |
35 |
35 |
35 |
Robust standard errors in parentheses; * significant at 10%; ** significant at 5%; *** significant at 1%.
Estimation method: Arellano-Bond GMM estimates (all variables in first differences).
Specification: always included, constant, lpop; polity_gt , and lagged Dep. var., once in col (1), twice in cols (2)—(5)).
In cols (1)-(5), instruments for lagged Dep. Var. is one additional lag of this variable.
In cols (4)-(5), party_frag is treated as endogenous with additional instruments: one lag of party_frag, maj, district, .and threshold..
AR(2) Arellano-Bond test for absence of second-order serial correlation.
Over-id: Sargan test of over-identifying restrictions, distributed as Chi2 with degrees of freedoms in parenthesis;
critical values at 5% confidence: 7.81 col (1), 3.84 cols (2)-(3), 12.59 cols (4)-(5).