Skip to main content
. 2014 Dec 12;12:70. doi: 10.1186/1478-4491-12-70

Table 3.

Difference-in-difference (DID) estimator, δ3 in the main analysis and robustness test using Year −3 as baseline

Year, t (1) Main analysis (2) Robustness test
First to third closest PS Fourth to sixth closest PS Seventh to ninth closest PS
−2 0.027a 0.019 0.025
(0.013) (0.013) (0.013)
−1 0.035b 0.033b 0.042b
(0.012) (0.012) (0.012)
0 0.021 0.006 0.018
(0.012) (0.013) (0.013)
1 0.057b 0.043b 0.051b
(0.015) (0.015 (0.016)
2 0.003 0.014 0.002
(0.025) (0.022) (0.028)
3 0.031 0.037 0.049
(0.027) (0.029) (0.026)
4 −0.019 −0.030 −0.017
(0.036) (0.038) (0.037)

Note: cluster robust standard errors in parentheses. Baseline: 3 years before merger approval by the regulator.

a P < 0.05; b P < 0.01.

PS, propensity score.