Table 6. The effect of tax progressivity on student achievement
|
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
(7) |
(8) | |
|
Income tax rate progressivity |
-0.141* |
-0.009 |
-0.086* |
-0.074+ |
-0.074 |
-0.133** |
-0.025 |
0.018 |
|
(0.058) |
(0.059) |
(0.042) |
(0.039) |
(0.057) |
(0.046) |
(0.050) |
(0.066) | |
|
GDP per capita (log) |
- |
1.297** |
1.113* |
1.031+ |
2.118+ |
0.001 |
1.399+ |
4.794** |
|
(0.226) |
(0.558) |
(0.523) |
(1.197) |
(0.258) |
(0.772) |
(1.534) | ||
|
Percentage secondary school |
- |
0.945** |
-0.601 |
-0.51 |
1.013 |
1.403** |
-0.067 |
0.637 |
|
attained among adults (log) |
(0.336) |
(0.496) |
(0.449) |
(1.043) |
(0.285) |
(0.671) |
(1.258) | |
|
Population size (log) |
- |
0.109 |
0.654 |
-0.632 |
0.857 |
0.005 |
4.553 |
-1.245 |
|
(0.085) |
(1.245) |
(1.178) |
(10.023) |
(0.061) |
(2.789) |
(10.999) | ||
|
trend |
- |
- |
- |
-0.006 (0.017) |
- |
- |
- |
- |
|
Country fixed effects |
No |
No |
Yes |
Yes |
Yes |
No |
Yes |
Yes |
|
Time fixed effects |
No |
Yes |
Yes |
No |
Yes |
Yes |
Yes |
Yes |
|
County specific trends |
No |
No |
No |
No |
Yes |
No |
No |
Yes |
|
Observations |
206 |
180 |
180 |
180 |
180 |
116 |
116 |
116 |
|
No of countries |
64 |
56 |
56 |
56 |
56 |
28 |
28 |
28 |
|
Sample |
All |
All |
All |
All |
All |
OECD |
OECD |
OECD |
|
R2 |
0.029 |
0.387 |
0.958 |
0.952 |
0.983 |
0.280 |
0.855 |
0.939 |
|
R2 (within) |
- |
- |
0.188 |
0.066 |
0.674 |
- |
0.290 |
0.700 |
Note. Absolute standard errors in parentheses, +, * and ** denote significance at 10, 5 and 1 percent level,
respectively.
Regarding OECD countries, there is a strong negative effect of tax progressivity (column (6))
when the model does not include country fixed effects, but the effect disappears when country
fixed effects are added. This finding is similar to the pattern observed for government
consumption spending in the OECD countries (columns (6) to (8) of Table 4). In Table 6, it
most possibly suggests that the conditional within country variation of tax progressivity is too
small to identify a statistically significant effect.
To sum up, we find that a more income redistributive state as measured by the progressivity of
the income tax system is negatively correlated with student performance in international tests
of Mathematics and Science. Again, this result holds for the world sample, and in tendency
for the OECD sample. It is in strong support of our hypothesis.
5.4. Generosity of the welfare system
The sizes of government consumption and welfare spending, expressed in percentage of GDP,
are commonly viewed as proxies for the generosity of redistributive activities by the
government. However, in principle, generosity of social transfers can be more directly
assessed when values per recipient of social benefits in place of per capita numbers are
employed. However, precise information on number of recipients is not easily available. Thus,
21