sions in which the GPA variable is excluded and the condition of common sup-
port has been relaxed. When these conditions are met, a year of higher studies
is found to generate returns of 8.2 per cent for males and 5.1 per cent for fe-
males.
Table 10 Wage earnings regression on individuals with registration in higher
studies.
Dependent variable: Log annual wage earnings in 2002.
|
Males |
Females | |||
|
Years of studies |
.102*** |
.082*** |
.091 *** |
.051*** |
|
(.0056) |
(.0072) |
(.0087) |
(.0108) | |
|
AE * yrs of studies |
- .020*** |
- .013* |
- .018 ** |
- .011 |
|
(.0059) |
(.0072) |
(.0072) |
(.0106) | |
|
Grade point average |
No |
Yes |
No |
Yes |
|
N |
10772 |
7511 |
12163 |
9025 |
|
Adj R2 __________ |
.088 |
.073 |
.055 |
.043 |
Note I: *** Significant at the 1 % level. ** at the 5 % level * at the 10 % level.
Note II: Explanatory variables; dummies of foreign born, father’s and mother’s education, children
living at home in 1999, transfers received in 1999 from UI, sick-leave, pensions, study grants and
year of exam 1991-1999.
There are no significant differences in the returns to university studies between
females with and without a prior AE. The coefficient for males, significant at a
ten per cent level, points to a 1.3 per cent lower payoff to a year in higher edu-
cation compared with individuals never registered in AE.
Table 11 presents estimates using binary variables for the amount of higher
studies, allowing the returns to be non-linear. The presentation again includes
two versions of the regression estimates. The GPA overlap condition is now set
in the first version while in the second version, the binary variables indicating
years of studies are replaced with 244 binary variables which represent educa-
tional directions as well as number of years of study as reported in SUN2002.
The parameters of main interest are those in the lower part, associated with
interaction variables between prior registration in AE and years of higher
education. In general, with two years or more of university studies, the returns
to education for AE individuals are not found to be different compared to indi-
viduals without AE. In contrast, AE individuals with less than two years of
36
IFAU - Does adult education at upper secondary level influence annual wage earnings?