to the full sample that includes the pre-1990 tests, although the effect of government size is
still insignificant in the model with time-specific fixed effects.
Table 7. Generosity of the welfare state: OECD countries
(1) |
(2) |
(3) |
(4) | |
GDP per capita (log) |
0.766 |
-0.156 |
0.975 |
1.029 |
(0.919) |
(1.340) |
(1.211) |
(0.963) | |
Percentage secondary school attained |
0.707 |
0.747 |
1.578+ |
0.357 |
among adults (log) |
(0.658) |
(0.832) |
(0.846) |
(0.668) |
Population size (log) |
2.969 |
2.919 |
5.581+ |
2.619 |
(2.437) |
(3.523) |
(2.885) |
(2.518) | |
Pension spending (log) |
-1.057* (0.502) |
- |
- |
- |
Active labor market policy |
- |
-0.544* (0.209) |
- |
- |
Unemployment spending (log) |
- |
- |
-0.112 (0.182) |
- |
Health care spending (log) |
- |
- |
- |
0.104 |
Share of elderly (log) |
-0.783 (0.964) |
- |
- |
-2.048* (0.868) |
Unemployment rates |
- |
-0.027 (0.032) |
0.000 |
- |
Country fixed effects |
yes |
yes |
yes |
Yes |
Time fixed effects |
yes |
yes |
yes |
Yes |
Observations |
110 |
110 |
113 |
110 |
Countries |
27 |
28 |
27 |
27 |
R2 |
0.8632 |
0.8687 |
0.8479 |
0.8544 |
R2 (within) |
0.3328 |
0.3995 |
0.3079 |
0.2902 |
F-test of joint significance |
5.5078 |
3.458 |
0.2524 |
3.1086 |
(p-value) |
0.006 |
0.0372 |
0.7776 |
0.051 |
Note. Absolute t-values in parentheses, +, * and ** denote significance at 10, 5 and 1 percent level, respectively.
It might also be argued that government size and the generosity of the welfare state are
proxies for educational expenditures, albeit some empirical evidence both from analyses
within and between countries indicates that educational expenditures have at most a minor
effect on student performance (e.g. Hanushek and Luque 2003).21 Similarly, welfare effects
might also proxy school (or teacher) quality - more generous governments in term of public
21 For contrasting evidence exploiting variation across Swiss states, see Fischer (2007). She finds that
educational spending exerts a decisive impact through teachers’ wages.
23