Table 8 Oster (2019) test for omitted...

Table 8

Oster (2019) test for omitted variable bias

Dependent variables	Controlled effect (⁠ $\hat{β} \hat{ˆ}$ ⁠)	Oster bounds (⁠ $\hat{β} \hat{ˆ}$ ⁠, $β^{*}$ ⁠)	Delta (δ)
Investment inefficiency	0.1145***	[0.1145, 0.0434]	1.047 > 1
Overinvestment	0.0213***	[0.0213, 0.0925]	1.052 > 1
Underinvestment	0.0649***	[0.0649, 0.0035]	1.068 > 1
Baseline controls	YES	YES	YES
Continent	YES	YES	YES
Excluded zero		YES

Dependent variables	Controlled effect ( $\hat{β} \hat{ˆ}$ )	Oster bounds ( $\hat{β} \hat{ˆ}$ , $β^{*}$ )	Delta (δ)
Investment inefficiency	0.1145***	[0.1145, 0.0434]	1.047 > 1
Overinvestment	0.0213***	[0.0213, 0.0925]	1.052 > 1
Underinvestment	0.0649***	[0.0649, 0.0035]	1.068 > 1
Baseline controls	YES	YES	YES
Continent	YES	YES	YES
Excluded zero		YES

Note(s): This table provides the Oster (2019) bounds for the key variable in the baseline analysis (Table 3). Here, δ indicates the extent of selection on unobserved variables relative to observed ones, while β* denotes the bias-adjusted coefficient, assuming δ = 1

Source(s): Authors own work

[ViewLarge]

Sharing Unavailable