Skip to main content

Table 4 Alternative estimates for public education spending per (FTE) student – tackling outliers

From: An empirical inquiry into the determinants of public education spending in Europe

SUR estimates for the model specifications:

Extended & GDP per capita

Extended & GDP per capita

Extended & GDP per capita

Extended & GDP per capita

Specification

(4.i)

(4.ii)

(7.i)

(7.ii)

Δlog (spending per student, FTE, at ISCED 0–1) [eq1]

    

 Δlog (ULC)

0.95***

0.88***

0.96***

0.88***

(4.61)

(3.93)

(4.65)

(3.94)

 Δlog (real GDP per capita)

0.55***

0.90***

0.56***

0.92***

(2.83)

(3.32)

(2.87)

(3.37)

 Δlog (pupil/teacher ratio) (−1)

−0.11

−0.09

−0.02

−0.03

(−0.82)

(−0.66)

(−0.16)

(−0.27)

 Δlog (share of capital expenditure) (−1)

0.05

0.05

0.05

0.05

(1.67)

(1.59)

(1.74)

(1.62)

 Δlog (share of teachers’ wages expenditure) (−1)

0.53***

0.36*

0.44**

0.31

(2.64)

(1.83)

(2.25)

(1.62)

 Constant

8.16**

2.78

8.03**

2.90

(2.53)

(0.75)

(2.49)

(0.78)

Δlog (spending per student, FTE, at ISCED 2–4) [eq2]

    

 Δlog (ULC)

0.47**

0.42*

0.48**

0.43*

(2.15)

(1.74)

(2.17)

(1.74)

 Δlog (real GDP per capita)

0.49**

0.82***

0.49**

0.83***

(2.38)

(2.76)

(2.40)

(2.78)

 Δlog (pupil/teacher ratio) (−1)

−0.36**

−0.26**

−0.31**

−0.23*

(−2.54)

(−1.78)

(−2.23)

(−1.65)

 Δlog (share of capital expenditure) (−1)

0.04

0.01

0.04

0.01

(1.07)

(0.35)

(1.08)

(0.35)

 Δlog (share of teachers’ wages expenditure) (−1)

0.54**

0.44**

0.49**

0.43**

(2.51)

(2.08)

(2.31)

(1.99)

 Constant

7.11**

−0.89

7.04**

−0.84

(2.08)

(−0.22)

(2.06)

(−0.21)

Δlog (spending per student, FTE, at ISCED 5–6) [eq3]

    

 Δlog (ULC)

0.86***

0.77***

0.86***

0.77***

(3.81)

(3.15)

(3.81)

(3.15)

 Δlog (real GDP per capita)

0.45**

0.72**

0.45**

0.72**

(2.12)

(2.46)

(2.12)

(2.46)

 Δlog (share of capital expenditure) (−1)

0.06

0.04

0.06

0.04

(1.60)

(1.22)

(1.60)

(1.22)

 Δlog (share of teachers’ wages expenditure) (−1)

0.53**

0.53**

0.54**

0.53**

(2.42)

(2.48)

(2.47)

(2.48)

 Constant

1.59

−0.45

1.59

−0.45

(0.45)

(−0.11)

(0.45)

(−0.11)

Δlog (general govt. spending per capita) [eq4]

    

 Δlog (ULC)

  

0.69***

0.53***

  

(6.30)

(4.67)

 Δlog (real GDP per capita)

  

0.43***

0.71***

  

(4.25)

(5.24)

 Constant

  

2.92*

3.96**

  

(1.70)

(2.08)

Observations

206

206

206

206

R 2 [eq1]

0.28

0.31

0.27

0.31

R 2 [eq2]

0.24

0.22

0.24

0.22

R 2 [eq3]

0.18

0.22

0.18

0.22

R 2 [eq4]

  

0.29

0.37

AIC

4567.29

4564.37

5799.84

5779.91

BIC

4723.70

4730.77

5999.51

5992.89

Country dummies for selected countries

Yes

-

Yes

-

Joint Wald-type test for dummies (at 5%)

eq2

 

eq2

 

Year dummies

-

Yes

-

Yes

Joint Wald-type test for dummies (at 5%)

 

eq1

 

eq1

Breusch-Pagan test of independence stat. (p-val.)

22.11 (0.00)

19.49 (0.00)

38.15 (0.00)

30.34 (0.00)

  1. Note: The R 2 is just a summary measure of the overall in-sample predictive power of the model; it is simply computed as 1-residual sum of squares/total sum of squared residuals. The null of the Breusch-Pagan tests is that the residuals across the estimated equations in the model are independent. For the joint Wald-type test, the table indicates the equation for which the null was rejected at the 5% level (meaning that dummies were jointly and statistically significant)
  2. Country and year dummies are not shown
  3. Standard errors are adjusted for small sample sizes. t statistics are given in parentheses, * p < 0.10, ** p < 0.05, *** p < 0.01