Table 3 presents some descriptive statistics on government consumption in the world sample,
and social expenditures for OECD countries, including its single components. The variance in
social expenditures is slightly higher than that for government consumption, both overall and
within countries. The within country variation, for which we identify the effects on student
achievement, constitutes 7-8 percent of the overall variance. Pension spending is the largest
component of social expenditures, followed by public health spending.
Table 3. Descriptive statistics of government consumption and social expenditures
Obser- |
Mean |
Standard |
Standard |
Minimum |
Maximum | |
General government consumption |
232 |
17.65 |
5.39 |
1.46 |
5.69 |
41.47 |
General government consumption |
124 |
18.90 |
4.22 |
1.05 |
10.08 |
29.62 |
Public sector social expenditures, |
124 |
19.62 |
5.61 |
1.62 |
2.8 |
32.5 |
Active labor market policy |
120 |
0.61 |
0.44 |
0.20 |
0 |
2.2 |
Public health spending, share of |
124 |
5.56 |
1.28 |
0.55 |
1.4 |
8.3 |
Family allowance spending, share |
124 |
1.90 |
1.08 |
0.31 |
0 |
4.1 |
Unemployment benefit spending, |
120 |
1.17 |
0.90 |
0.44 |
0 |
4.4 |
Pension spending, |
124 |
6.38 |
2.76 |
0.76 |
0.6 |
12.8 |
Housing spending, |
102 |
0.42 |
0.39 |
0.17 |
0 |
1.8 |
Other social spending, |
99 |
4.00 |
1.38 |
0.64 |
1.50 |
8.90 |
The third measure of the generosity of the welfare state we employ is an index of income tax
rate progressivity developed by the Fraser Institute (Gwartney and Lawson, 2002). The index
constitutes an income-bracket adjusted marginal tax rate levied in the highest income bracket
in one country, adjusted for the lowest income threshold for this income bracket. The
redistributive impact of a given tax rate depends on the financial threshold from which on the
rate applies. The top tax rate index of the Fraser Institute is therefore adjusted for threshold
effects to facilitate comparability of the marginal top income tax rate across countries and
14