Indonesian Higher Education 135
TABLE 5.10
Non-Linear Decomposition of Higher Education Participation with Selectivity Bias
Correction between Public School and Private SchoolsPrivate NR Private Islam Private Christian
Higher Education Participation 0.172 –0.136 0.249
Differential
Contributions from school choice
differences in:
INCOME 0.114 0.067 0.042*
(0.005) (0.004) (0.004)
SCHOOL 0.000 0.002 0.001
(0.000) (0.002) (0.002)
PERSONAL –0.010* 0.016 –0.025
(0.004) (0.006) (0.024)
FAMILY 0.002 0.000 –0.010*
(0.003) (0.002) (0.005)
ABILITY 0.011 0.008 –0.007
(0.003) (0.003) (0.003)
PARENTS 0.033** 0.037 –0.002
(0.004) (0.004) (0.003)
N 1 1,717 1 1,398 1 1,176
Notes: Standard errors are in parenthesis and heteroscedasticity consistent. Significance at 10 per cent
level, Significance at 5 per cent level, and Significance at 1 per cent level. Dummies representing
the province where individual was lived and selectivity variables were also included in the model.
The smallest contributor to the gap is personal characteristics and, in fact,
this variable contributes negatively to the gap. The higher number of non-
Muslim students from wealthy family in private secular school than public
school could be one of the reasons for this gap is negative.
The family background (FAMILY), and school quality (SCHOOL) do
not play a major role in explaining public and private secular differences
in higher education participation. FAMILY contributes 0.002 to the gap
whereas SCHOOL only contributes a little bit more than –0.000. SIBLING
and WORKSMA have positive contribution whereas URBAN contributes
negatively to the gap.
The disparity of probability of higher education participation between
public and private Islam school group is smaller than private secular gap,
but still quite large or 0.172. The non-linear decomposition results between
public and private Islam group are in column 3 of Tables 5.9 and 5.10.
There are four determinants that statistically significantly contribute to the