Table 3. Estimation results for RD...

Table 3.

Estimation results for RD analysis

Depended variable: winning dummy variable	Model 1: general competitive bidding			Model 2: designated bidding			Model 3: designated bidding
Depended variable: winning dummy variable	Coef.	Std. Err.	Robust P > z	Coef.	Std. Err.	Robust P > z	Coef.	Std. err.	Robust P > z
Estimated discontinuity (robust)	−0.076	0.077	0.322	−0.454	0.132	0.001	−0.444	0.134	0.001
Cutoff c = 0.95	Left of c	Right of c		Left of c	Right of c		Left of c	Right of c
Number of obs.	1,552	2,692		2,218	7,001		2,218	7,001
Effective number of obs.	557	973		1,139	1,486		1,451	2,781
Order est. (p)	1	1		1	1		2	2
Order bias (q)	2	2		2	2		3	3
BW est. (h)	0.015	0.015		0.012	0.012		0.020	0.020
BW bias (b)	0.019	0.019		0.016	0.016		0.025	0.025
Rho (h/b)	0.810	0.810		0.759	0.759		0.821	0.821
Number of clusters	21	21		21	21		21	21
Number of obs.	4,244			9,219			9,219
BW type	mserd
Kernel	Triangular
VCE method	Cluster (municipality dummy)

Depended variable: winning dummy variable	Model 1: general competitive bidding			Model 2: designated bidding			Model 3: designated bidding
Depended variable: winning dummy variable	Coef.	Std. Err.	Robust P > z	Coef.	Std. Err.	Robust P > z	Coef.	Std. err.	Robust P > z
Estimated discontinuity (robust)	−0.076	0.077	0.322	−0.454	0.132	0.001	−0.444	0.134	0.001
Cutoff c = 0.95	Left of c	Right of c		Left of c	Right of c		Left of c	Right of c
Number of obs.	1,552	2,692		2,218	7,001		2,218	7,001
Effective number of obs.	557	973		1,139	1,486		1,451	2,781
Order est. (p)	1	1		1	1		2	2
Order bias (q)	2	2		2	2		3	3
BW est. (h)	0.015	0.015		0.012	0.012		0.020	0.020
BW bias (b)	0.019	0.019		0.016	0.016		0.025	0.025
Rho (h/b)	0.810	0.810		0.759	0.759		0.821	0.821
Number of clusters	21	21		21	21		21	21
Number of obs.	4,244			9,219			9,219
BW type	mserd
Kernel	Triangular
VCE method	Cluster (municipality dummy)

Note(s):

The coefficients, standard errors and significance levels are robust estimates. One bidding process (with three bids) was excluded due to the lack of municipality dummies, reducing the number of observations. The polynomial order is set to one for estimation in Models 1 (general competitive bidding) and 2 (designated bidding). Model 3 extends the polynomial regression order to two for designated bidding to confirm the robustness of Model 2’s results

Source(s): Author’s own work

[ViewLarge]

Sharing Unavailable